18
2 State of the art
HSDT. However, in industrial applications single layer FSDT models are most common.
They are readily available in commercial finite element software and enable to analyze
large structures at low computational effort. While these models are effective in representing symmetric in-plane damage mechanism, they are limited in their capability to
represent out-of-plane sandwich failure modes, like core crushing, face wrinkling and
delamination. This can be overcome by implementing 3D-continuum models. By modelling the core with solid elements or quasi-solid elements in case of axial symmetry, various out-of-plane damage mechanisms can be simulated using effective homogenized
mechanical core properties. Such a core is often coupled with 2D-shell elements for the
faces, leading to the well-established Shell-Solid-Shell approach. The increased damage
modeling capability of this approach, if compared to the simplified and 2D-models, results in increased computational effort. However, due to the homogenization of the core
no buckling of the honeycomb cell walls can be represented. This can be achieved using
detailed models, where the actual cellular core geometry is modelled. This is required to
simulate the large local deformation of the core cell walls accurately. These meso-scale
models require high computational effort, yet they have seen increasing applications in
the recent past, due to increasing computational capabilities. Figure 16 illustrates the
four introduced categories, while the relative damage modeling capabilities and computational effort are indicated using an arrow. In the following, the Finite Element Method
is briefly introduced as fundamental tool in the present work before a literature survey
on computational models for honeycomb sandwich panel joints is given.
Simplified models
2D shell and plate
models
3D-Continuum models
Detailed models
Analytical equations
Global approximation, discrete layer
e.g. Shell-Solid-Shell
Meso/Micro scale
Increasing computational effort
Increasing damage modeling capabilities
Figure 16 Categories of computational models for sandwich structures [See14]
2 State of the art
HSDT. However, in industrial applications single layer FSDT models are most common.
They are readily available in commercial finite element software and enable to analyze
large structures at low computational effort. While these models are effective in representing symmetric in-plane damage mechanism, they are limited in their capability to
represent out-of-plane sandwich failure modes, like core crushing, face wrinkling and
delamination. This can be overcome by implementing 3D-continuum models. By modelling the core with solid elements or quasi-solid elements in case of axial symmetry, various out-of-plane damage mechanisms can be simulated using effective homogenized
mechanical core properties. Such a core is often coupled with 2D-shell elements for the
faces, leading to the well-established Shell-Solid-Shell approach. The increased damage
modeling capability of this approach, if compared to the simplified and 2D-models, results in increased computational effort. However, due to the homogenization of the core
no buckling of the honeycomb cell walls can be represented. This can be achieved using
detailed models, where the actual cellular core geometry is modelled. This is required to
simulate the large local deformation of the core cell walls accurately. These meso-scale
models require high computational effort, yet they have seen increasing applications in
the recent past, due to increasing computational capabilities. Figure 16 illustrates the
four introduced categories, while the relative damage modeling capabilities and computational effort are indicated using an arrow. In the following, the Finite Element Method
is briefly introduced as fundamental tool in the present work before a literature survey
on computational models for honeycomb sandwich panel joints is given.
Simplified models
2D shell and plate
models
3D-Continuum models
Detailed models
Analytical equations
Global approximation, discrete layer
e.g. Shell-Solid-Shell
Meso/Micro scale
Increasing computational effort
Increasing damage modeling capabilities
Figure 16 Categories of computational models for sandwich structures [See14]