Optimale Auslegung versteifter Faserverbund-Flaechentragwerke unter Beul-und Eigenfrequenzrestriktionen

In the field of lightweight construction, one often employs stiffened shell structures mades of fibre-composite materials as these structures display a high specific ratio of stiffness and strength to weight. However, a design with respect to weight has to consider the fact that these structures are endangered by buckling when subject to inplane compression and shear loads. In addition to the eigenvibration behaviour, the critical bucking load is therefore an essential dimensioning criterion for lightweight structures. In structural optimization special procedures are provided by means of which the variable structural parameters ply-angles, ply-thicknesses, stiffener cross-sections, and distance between the stiffeners can most suitably be adapted to given demands like low weight, high stiffness, high buckling load, etc. For the computation of the structural responses, an FEM-program is described that has been adjusted to the type of analysis (analysis of buckling and eigenvibrations) and to the component geometry. In contrast to commercial programs, this system has significant advantages regarding computational time, and it is also able to realize problem-specific requirements for the optimization as, for instance, analytical sensitivity analysis for eigenvalue problems as well as a special eigenvalue constraint formulation. In recognizing the eigenmodes corresponding with the eigenvalues during iteration, the latter formulation complies with the demand for continuity of the optimization algorithms with respect to constraint gradients. An axial stiffened CFRP-shell is used as an important application example to test the developed models and procedures. The structural modelling for a design optimized in terms of weight and buckling load and for a standard design is verified with experimental investigations. (orig.)Available from TIB Hannover: RR 2225(10-06) / FIZ - Fachinformationszzentrum Karlsruhe / TIB - Technische InformationsbibliothekSIGLEDEGerman