@inproceedings{KavrakovTimmlerMorgenthal,
author = {Kavrakov, Igor and Timmler, Hans-Georg and Morgenthal, Guido},
title = {STRUCTURAL OPTIMIZATION USING THE ENERGY METHOD WITH INTEGRAL MATERIAL BEHAVIOUR},
series = {Digital Proceedings, International Conference on the Applications of Computer Science and Mathematics in Architecture and Civil Engineering : July 20 - 22 2015, Bauhaus-University Weimar},
booktitle = {Digital Proceedings, International Conference on the Applications of Computer Science and Mathematics in Architecture and Civil Engineering : July 20 - 22 2015, Bauhaus-University Weimar},
editor = {G{\"u}rlebeck, Klaus and Lahmer, Tom},
organization = {Bauhaus-Universit{\"a}t Weimar},
issn = {1611-4086},
doi = {10.25643/bauhaus-universitaet.2806},
url = {http://nbn-resolving.de/urn:nbn:de:gbv:wim2-20170314-28065},
pages = {6},
abstract = {With the advances of the computer technology, structural optimization has become a prominent field in structural engineering. In this study an unconventional approach of structural optimization is presented which utilize the Energy method with Integral Material behaviour (EIM), based on the Lagrange's principle of minimum potential energy. The equilibrium condition with the EIM, as an alternative method for nonlinear analysis, is secured through minimization of the potential energy as an optimization problem. Imposing this problem as an additional constraint on a higher cost function of a structural property, a bilevel programming problem is formulated. The nested strategy of solution of the bilevel problem is used, treating the energy and the upper objective function as separate optimization problems. Utilizing the convexity of the potential energy, gradient based algorithms are employed for its minimization and the upper cost function is minimized using the gradient free algorithms, due to its unknown properties. Two practical examples are considered in order to prove the efficiency of the method. The first one presents a sizing problem of I steel section within encased composite cross section, utilizing the material nonlinearity. The second one is a discrete shape optimization of a steel truss bridge, which is compared to a previous study based on the Finite Element Method.},
subject = {Angewandte Informatik},
language = {en}
}
@inproceedings{RaueTimmlerSchroeter,
author = {Raue, Erich and Timmler, Hans-Georg and Schr{\"o}ter, Hendrik},
title = {NON-LINEAR ANALYSIS OF SHELLS OF REVOLUTION USING MATHEMATICAL OPTIMISATION},
editor = {G{\"u}rlebeck, Klaus and K{\"o}nke, Carsten},
organization = {Bauhaus-Universit{\"a}t Weimar},
issn = {1611-4086},
doi = {10.25643/bauhaus-universitaet.2881},
url = {http://nbn-resolving.de/urn:nbn:de:gbv:wim2-20170314-28818},
pages = {12},
abstract = {In the paper presented, reinforced concrete shells of revolution are analyzed in both meridional and circumferential directions. Taking into account the physical non-linearity of the material, the internal forces and the deflections of the shell as well as the strain distribution at the cross-sections are calculated. The behavior of concrete under compression is described by linear and non-linear stress-strain relations. The description of the behavior of concrete under tension must account for tension stiffening effects. A tri-linear function is used to formulate the material law of reinforcement. The problem cannot be solved analytically due to the physical non-linearity. Thus a numerical solution is formulated by means of the LAGRANGE Principle of the minimum of the total potential energy. The kinematically admissible field of deformation is defined by the displacements u in the meridional and w in the radial direction. These displacements must satisfy the equations of compatibility and the kinematical boundary conditions of the shell. The strains are linearly distributed across the wall thickness. The strain energy depends on the specific of the material behavior. Using integral formulations of the material law [1], the strain energy of each part of the cross-section is defined as a function of the strains at the boundaries of the cross-sections. The shell is discretised in the meridional direction. Various methods of numerical differentiation and numerical integration are applied in order to determine the deformations and the strain energy. The unknown displacements u and w are calculated by a non-restricted extremum problem based on the minimum of the total potential energy. From mathematical point of view, the objective function is a convex function, thus the minimum can be determined without difficulty. The advantage of this formulation is that unlike non-linear methods with path-following algorithms the calculation does not have to account for changing stiffness and load increments. All iterations necessary to find the solution are integrated into the "Solver". The model presented provides many ways of investigating the influence of various material parameters on the stresses and deformations of the entire shell structure.},
subject = {Angewandte Informatik},
language = {en}
}