Page 51 - Fundamentals of Computational Geoscience Numerical Methods and Algorithms

Chapter 3
A Consistent Point-Searching Interpolation

Algorithm for Simulating Coupled Problems
between Deformation, Pore-Fluid Flow, Heat
Transfer and Mass Transport Processes
in Hydrothemal Systems

Over the past decade or so, many commercial computational codes have become
available for solving a great number of practical problems in both scientific and
engineering fields. Primary advantages of using commercial computational codes
are: (1) built-in pre-processing and post-processing tools make it very easy and
attractive to prepare, input and output data which are essential in a numerical analy-
sis; (2) provision of movie/animation functions enables numerical results, the treat-
ment of which is often a cumbersome and tedious task, to be visualised via clear and
colourful images; (3) detailed benchmark solutions and documentation as well as
many embedded robust solution algorithms allow the codes to be used more easily,
correctly, effectively and efficiently for solving a wide range of practical problems.
However, the main disadvantage of using commercial computational codes is that
each code is often designed, within a certain limit, for solving some particular kinds
of practical problems. This disadvantage becomes more and more obvious because
the ever-increasing competitiveness in the world economy requires us to deal with
more and more complicated and complex geoscience problems, which are encoun-
tered and not solved in the field of contemporary computational geoscience. There
are three basic ways to overcome the above difficulties. The first is to develop some
new commercial computational codes, which is time consuming and often not cost-
effective for numerical analysts and consultants. The second is to extend an existing
commercial computational code, which is usually impossible because the source
code is often not available for the code users. The third is to use several existing
commercial computational codes in combination. This requires development of a
data translation tool to transfer data necessary between each of the codes to be used.
Compared with the difficulties encountered in the first two approaches, the third one
is more competitive for most numerical analysts and consultants.
Our first successful example in the practice of using commercial computa-
tional codes in a combination manner was to optimize structural topologies under
either static or dynamic conditions using the commercial code STRAND6 (G+D
Computing 1991) and a home-made code GEMDYN. As a result, a generalized
evolutionary method for numerical topological optimization of structures has been
developed and many interesting numerical results have been produced (Zhao et al.
1996d, e, 1997b, c, d, 1998c, d). To extend further the idea of using commercial

C. Zhao et al., Fundamentals of Computational Geoscience, 37
Lecture Notes in Earth Sciences 122, DOI 10.1007/978-3-540-89743-9 3,
C Springer-Verlag Berlin Heidelberg 2009