Keywords: identification, inverse analysis, microplane model, multi-objective optimization, evolutionary algorithms, parallel processing.

Concrete is one of the most frequently used materials in civil engineering. Nevertheless, as a highly heterogeneous system, it shows very complex non-linear behavior, which is extremely difficult to describe by a sound constitutive law. As a consequence, a numerical simulation of response of complex concrete structures still remains a very challenging and demanding topic in engineering computational modeling.

One of the most promising approaches to modeling of concrete behavior is based on the microplane concept [1]. It is a fully three-dimensional material law that incorporates tensional and compressive softening, damage of the material, supports different combinations of loading, unloading and cyclic loading along with the development of damage-induced anisotropy of the material. The major disadvantages of this model are, however, a large number of phenomenological material parameters and a high computational cost associated with structural analysis even in a parallel environment.

This year, a procedure based on artificial neural networks (ANNs) [2] for the microplane parameter identification that is able to identify reliably all microplane parameters was developed. In particular, an artificial neural network was used to estimate required parameters. As the training procedure, the genetic algorithm-based method was used. However, the drawback of this methodology is a high computational cost of the identification algorithm. A suite of thirty uniaxial compression tests consumes approximately 25 days on a single processor PC with the Pentium IV 3400 MHz processor and 3 GB RAM.

In this contribution, the problem is solved by implementing a parallel multi-objective procedure. The numerical analysis is implemented using the OOFEM - free finite element code with object oriented architecture. The optimization procedure utilizes the global parallel model [3]. More specifically, the program is divided into an optimization part and an analysis part. In this way it is implemented in the cluster of PCs. As an optimization algorithm, the method called SADE [4] is used. Management of several objectives is utilized by the average ranking procedure [5]. Errors among experimental and computed stress-strain curves are used as objectives.

1

Z.P. Bazant, F.C. Caner, I. Carol, M.D. Adley, S.A. Akers, "Microplane model M4 for concrete. Part I: Formulation with work-conjugate deviatoric stress. doi:10.1061/(ASCE)0733-9399(2000)126:9(944); Part II: Algorithm and calibration". doi:10.1061/(ASCE)0733-9399(2000)126:9(954), Journal of Engineering Mechanics-ASCE, 126, 944-953, 954-961, 2000.

2

A. Kucerová, M. Lepš, J. Zeman, "Back Analysis of Microplane Model Parameters Using Soft Computing Methods", Computer Assisted Mechanics in Engineering Sciences, 14(2), 2007.

3

E. Cantú-Paz, Efficient and Accurate Parallel Genetic Algorithms, Kluwer Academic Publishers, 2001.

4

O. Hrstka, A. Kucerová, "Improvements of real coded genetic algorithms based on differential operators preventing the premature convergence", Advances in Engineering Software, 35, 237-246, 2004. doi:10.1016/S0965-9978(03)00113-3

5

P.J. Bentley and J.P. Wakefield, "An Analysis of Multiobjective optimization within Genetic Algorithms", Technical Report ENGPJB96, University of Huddersfield, UK, 1996.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £62 +P&P)