652
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
3.1.3
Typical issues related to constitutive modelling
In real soils and rocks, their constitutive behaviors may be non-
unique arising from non-associated plasticity and strain-
softening, undrained and unsaturated behaviour.
3.1.4
Examples of uncertainties
Assumptions are normally made in numerical modeling because
of the lack of soil data, the lack of spatial variation of soil
properties, the lack of information of shaking information
during an earthquake, and the lack of information of future
developments around the project to be designed. The actual
construction may also deviate from the original design.
3.1.5
Specific software and hardware issues
Users are of no control of the specific implementations made by
the developers of the software. The software may depend on
operating system and configuration of computer. ‘Bugs’
(programming flaws in the application software) may exist and
they will only appear when certain constraints or limitations of
the software are encountered. Issues may also come from the
specific implementations of models (for example rounding-off
the corners of the Mohr-Coulomb failure criterion), the iterative
solvers and their numerical solution tolerances, and the parallel
solvers (solution differences depending on the number of
threads or cores being used).
3.1.6
Examples of misinterpretation of results
Some users of FEM may miss-use the safety factors,
misinterpret structural behaviour (if the structure is too much
simplified), overlook essential details (in particular complex 3D
models), and possess insufficient knowledge and understanding
of the modelling software being used.
3.2
Methods of validation
Brinkgreve and Engin (2013) gave the following examples of
validating FEM models and methods: (i) analytical solutions of
elasticity problems, plasticity problems, constitutive models,
dynamic problems, bearing capacity solutions, solutions of flow
and coupled problems, (ii) limit equilibrium solutions for global
safety factors or bearing capacities, (iii) upper and lower bound
solutions (limit analysis), and (iv) benchmarks. The following
aspects of validation were discussed in full details by
Brinkgreve and Engin (2013): (i) Validation of constitutive
models and parameters, (ii) Validation of model boundaries, (iii)
Validation of initial conditions, (iv) Validation of (the accuracy
of) results, (v) Benchmarking, and (vi) Checklists.
This paper is a very useful reference and is a must read for
serious practitioners of numerical analysis and young engineers.
4 CONCLUSION
In this brief report, we have summarized a total of 52 papers
submitted to the area of TC103 “Numerical Methods”. Instead
of summarized each paper, we have provided an overall view of
where these papers were from. A master table is given for all 52
papers in terms of the types of numerical methods empolyed by
different authors together with the full references given in the
end of the paper (paper number in alphabetic order). The
numerical methods used include finite element method (FEM),
finite difference method (FDM), material point method (MPM),
smoothed particle hydrodynamics (SPH), neural network (NN),
genetic algorithm (AG), and finite volume method (FVM). The
failure models used in studies include Mohr-Coulomb failure
criterion, Drucker-Prager plastic potential, Cam clay model,
Matsuoka-Nakai failure model, and Hoek-Brown failure
criterion. In terms applications, these numerical analyses have
been applied to model piles, tunnels, retaining walls, slopes,
levees, tailings impoundment, and breakwaters. Experiments
used in calibrating these numerical models can be classified as
centrifuge tests and 1-g laboratory tests. Finally, the interesting
report on the vailidation of finite element method by Brinkgreve
and Engin (2013) was summarized briefly.
In terms of future challenges, we expect more research work
considering the coupling between FEM and FDM (this was also
considered by Javadi et al., 2013). It is because FEM is good for
far field linear response whereas FDM (as such FLAC) better
suits the near field large deformation and failure. In terms of
multi-physics problems, we expect future developments in
numerical methods for coupling between thermal, electrical,
fluid, solids and chemical problems (like petroleum or thermal
energy extraction problems). Multi-physics problems are
commonly encountered in petroleum mining and geothermal
energy extraction problems. In fact, Yoneda (2013) was
motivated by the extract of methane hydrate (a new source of
energy fuel in Japan). In terms of multi-scale approach, we
expect more refined model for coupling of microscopic
behavior (such as grain rotations and sliding) to macroscopic
behavior (such as plastic yielding) in soils and coupling of
microscopic cracking process to macroscopic damage (at
continuum scale) and material degradation in rocks (Borja,
2011). This is sometimes referred as micro-meso-macro
modeling. For fluid-solid coupling problems, lattice Boltzman
model can be useful in fluid flow simulation modeling. In rock
blasting or fragmentation, computational simulations of
dynamic fragmentation of rocks is another area of active
research.
A typical example of multi-physics and multi-scale problems
in geomechanics is the so-called carbon capture and storage
(CCS). Carbon capture and storage (CCS) (or carbon capture
and sequestration), is the process of capturing waste carbon
dioxide (CO
2
) from large point sources, such as fossil fuel
power plants, transporting it to a storage site, and depositing it
where it will not enter the atmosphere, normally an
underground geological formation (like man-made rock
caverns). The aim is to prevent the release of large quantities of
CO
2
into the atmosphere (from fossil fuel use in power
generation and other industries). Another related problem is
nuclear power waste storage in underground rock layers.
Coupling effect of the thermo-hydro-mechanical responses in
rock becomes very important.
Another area that needs the use of accurate numerical
modeling is related to geohazards and geo-disasters. These
disasters include ground and basin amplification of earthquake
shaking, slope failure and landslides, and tsunami and storm
surge-induced failure of levees and breakwaters.
5 ACKNOWLEDGEMENTS
This paper was written when KTC was fully supported by a
grant from the Research Grants Council of the Hong Kong SAR
Government through Project No. PolyU 5002/08P. The author
is grateful to Prof. Jidong Zhao of HKUST and Prof. Richard
Wan of University of Calgary for helpful discussions.
6 REFERENCES OF 52 PAPERS
Balakumar V. Huang M., Oh E. and Balasubramaniam A.S. 2013.
Equivalent pier theory for piled raft design.
Proceedings of the
18th International Conference on Soil Mechanics and
Geotechnical Engineering, Paris 2013
.
Bennani Y., Soyez L. and Freitag N. 2013. Interprétation d’essais
d’extraction de renforcements métalliques haute adhérence dans
un massif en Terre Armée® soumis à un chargement dynamique
cyclique.
Proceedings of the 18th International Conference on
Soil Mechanics and Geotechnical Engineering, Paris 2013
.
Biru A.and Benz T. 2013. On non-coaxial stress-dilantancy theories.
Proceedings of the 18th International Conference on Soil
Mechanics and Geotechnical Engineering, Paris 2013
.
