| dc.description.abstract | The structural behavior of thin-walled aluminum extrusions, with and without aluminum foam filler, subjected to oblique loading was studied through experiments and numerical analyses. Extrusions similar to the ones tested here have high energy-absorbing capability when subjected to pure axial loading. The present study shows, however, that introduction of only a small load angle decreases the energy absorption considerably. Two approaches to increase the energy absorption were considered: (1) increasing the wall thickness, and (2) introducing aluminum foam filler. Other factors that influence the energy absorption are of course the geometry of the investigated component, and the strain hardening of the extrusion material. To limit the number of parameters in the present project, a component of same shape and length was considered, while various combinations of load angle, cross-sectional thickness, temper, and foam density were tested. The main response parameters of interest were in addition to the energy absorption, the peak load and the force-displacement behavior.
The experimental results were compared with the recommendations of Eurocode 9, which proved to be conservative. In this context, the interaction between moment and axial forces was also examined for a distorted cross section. Furthermore, based on the experimental program, two ways of increasing the energy absorption were evaluated, i.e. an increase in wall thickness and introduction of foam filler. In terms of improved energy absorption for the same weight, increasing the wall thickness was more effective than using aluminum foam filler tested in the present project.
Material tests on the extrusion material and aluminum foam were carried out to provide the stress-strain characteristics of the material, and for calibration of proper material models for numerical investigations. In order to model the foam-filled extrusions, an existing material model for aluminum foam proposed by Deshpande and Fleck [1], where the hydrostatic stress is incorporated in the yield criterion, was evaluated and implemented in LS-DYNA [2] as a user subroutine. Additionally, statistical variation of density and two simple fracture criteria were included in the model. The implemented model was tested thoroughly and verified by comparing the results to experimental tests on aluminum foam available in the literature.
Numerical analyses of obliquely loaded extrusions with and without foam filler were carried out in LS-DYNA to validate a finite element model. An existing model (no. 103 in LS-DYNA) was used for the extrusion material, while the foam was modeled with the implemented Deshpande-Fleck model. The finite element model was able to predict the experimental results with reasonably accuracy. | nb_NO |
