To be competitive, more and more buildings have to be developed and constructed where technical and mechanical limits are maxed out. Today focus no longer lies purely on the functionality of the structure, but also on its architectural possibilities. Landmarks with outstanding features, such as their height, slenderness or spans are being built. Within the chosen context, the optimum design of a structure can be determined by use of the so-called structural optimization. The most commonly used formulation of the structural optimization problem is the conventional nested analysis and design approach (in short: NAND) where only design variables, like cross sectional areas or nodal coordinates, are treated as independent optimization variables. The structural variables, like nodal displacements, stresses or internal forces, are treated as dependent variables, i. e. they are defined as implicit functions of the design variables. This approach proved to be very efficient, particularly for structures with linear structural behaviour or multiple load cases. However, in most of the practical optimization problems nonlinear effects have to be considered. In these cases, the structural analysis and the sensitivity analysis of problems defined by the NAND-formulation is very expensive. To increase the efficiency of the optimization process of non-linear structures it could be of advantage to use a different formulation for the description of the problem, e. g. the so-called simultaneous analysis and design approach (in short: SAND), where in addition to the design variables certain structural variables are treated as independent optimization variables. Thus, the equations of the structural analysis can be treated as equality constraints and no sensitivity analysis has to be performed. This thesis has the objective to evaluate the proposed alternative optimization formulation SAND and compare it with the conventional NAND-formulation by solving various problems: - Strain-energy-minimization of a two-bar-truss - Weight minimization of a ten-bar-truss Thereby, the performance, the quality of the optimization and the robustness of the approaches SAND and NAND are evaluated. In addition, the advantages and disadvantages during the implementation in MATLAB are being discussed. Thus, the potential of the alternative SAND-optimization for complex practical applications, the combinability with new algorithms and the ability of implementation into existing software can be assessed.     «

To be competitive, more and more buildings have to be developed and constructed where technical and mechanical limits are maxed out. Today focus no longer lies purely on the functionality of the structure, but also on its architectural possibilities. Landmarks with outstanding features, such as their height, slenderness or spans are being built. Within the chosen context, the optimum design of a structure can be determined by use of the so-called structural optimization. The most commonly used f...     »