TY - JOUR
T1 - Computed tomography-based modeling of structured polymers
AU - Wismans, J.G.F.
AU - van Dommelen, J.A.W.
AU - Govaert, L.E.
AU - Meijer, H.E.H.
AU - van Rietbergen, B.
PY - 2009
Y1 - 2009
N2 - A hybrid numerical-experimental approach is proposed to characterize the macroscopic mechanical behavior of structured polymers. The method is based on capturing the details of the material’s microstructure using 3D X-ray Computed Tomography (CT). By employing segmentation and voxel-conversion, the reconstructed volume is automatically converted into a finite element model that is subsequently used for mechanical analyses. The approach is demonstrated on a 2D polycarbonate honeycomb. An ideal representative volume element (RVE), with a volume equivalent to the volume of the real X-ray CT-based model, is used to determine the dependence of the macroscopic response of the structure on intrinsic material behavior, strain rate and cell wall thickness. A nonlinear elasto-viscoplastic constitutive model is used to describe the intrinsic behavior of the polycarbonate base material and a comparison with a hyper-elastic material model reveals that local plastic deformation significantly influences the macroscopic behavior. A cubic relation between the stiffness of the structure and cell wall thickness is found, whereas the strain rate has a minor influence. The ideal RVE shows a different response compared to the real X-ray CT-based model due to local variations of the cell wall thickness in the latter, causing non-homogeneous deformations. In addition to the geometric imperfections, jagged edges, as a consequence of voxel conversion, contribute to this local variation in cell wall thickness.
AB - A hybrid numerical-experimental approach is proposed to characterize the macroscopic mechanical behavior of structured polymers. The method is based on capturing the details of the material’s microstructure using 3D X-ray Computed Tomography (CT). By employing segmentation and voxel-conversion, the reconstructed volume is automatically converted into a finite element model that is subsequently used for mechanical analyses. The approach is demonstrated on a 2D polycarbonate honeycomb. An ideal representative volume element (RVE), with a volume equivalent to the volume of the real X-ray CT-based model, is used to determine the dependence of the macroscopic response of the structure on intrinsic material behavior, strain rate and cell wall thickness. A nonlinear elasto-viscoplastic constitutive model is used to describe the intrinsic behavior of the polycarbonate base material and a comparison with a hyper-elastic material model reveals that local plastic deformation significantly influences the macroscopic behavior. A cubic relation between the stiffness of the structure and cell wall thickness is found, whereas the strain rate has a minor influence. The ideal RVE shows a different response compared to the real X-ray CT-based model due to local variations of the cell wall thickness in the latter, causing non-homogeneous deformations. In addition to the geometric imperfections, jagged edges, as a consequence of voxel conversion, contribute to this local variation in cell wall thickness.
U2 - 10.1177/0021955X08100045
DO - 10.1177/0021955X08100045
M3 - Article
SN - 0021-955X
VL - 45
SP - 157
EP - 179
JO - Journal of Cellular Plastics
JF - Journal of Cellular Plastics
IS - 2
ER -