5.1 Panel flexure
79
Figure 60 with the help of pictograms. The experimental results are characterized by little scatter in terms of bending stiffness (below 5%). The scatter in terms of strength is
slightly more pronounced (up to 10%). An exception is specimen B19-L101-03, which is
an outlier in terms of strength due to premature failure in form of local core indentation,
while the other two tested specimens of the same configuration indicate face sheet failure. The obtained test results generally reflect the expectations and thus appear to be
plausible. Thin panels with comparably weak face sheets (B10, B07) fail due to face sheet
rupture, while the B26 and B19 panels fail due to local core indentation, unless loading
plates were applied to prevent this. Core shear failure is only evident with strong face
sheets and in W-orientation where the core shear strength is the lowest.
5.1.2 Numerical analysis
Due to the importance of bending tests in sandwich construction, there are several numerical studies on flexural bending of honeycomb sandwich beams available in the literature. These studies can be subdivided according to the applied core modelling approach. Heimbs [Hei06] and Zinno et al. [Zin11] model flexural tests of Nomex honeycomb sandwich beams using 3D-continuum elements for the core. Their models are intended to enable validation of existing material properties, while the focus is on the
global panel response. Other models are targeted towards localized effects such as indentation or wrinkling. They therefore apply detailed meso-models. Giglio et al. [Gig12]
investigated indentation in case of three point bending of a Nomex core with aluminum
face sheets. They conclude that the friction parameter between load cylinder and face
sheets has significant impact on the indentation pattern of the simulation. They suggest
a friction parameter between 0 and 0.3. Staal [Sta06] studied face wrinkling of Nomex
honeycomb panels with GFRP face sheets using detailed meso models and 3D-continuum models. He reports that a meso-model has no benefits over 3D-continuum models
for the prediction of face wrinkling. In the present work all ten tested configurations are
implemented in a virtual testing framework. The objective is to validate and benchmark
previously derived modelling approaches and material properties. Therefore, the bending tests are implemented at meso-scale as well as using 3D-continuum elements for the
core. The experimentally observed failure modes enable to validate and benchmark the
model performance regarding the out of plane compressive and transverse shear
strength of the core models. In addition, the compressive strength of the face sheets is
calibrated and the implemented modelling approach for the face to core bond is evaluated. In the following, the implemented models are described in detail before the simulation results are presented in comparison to the experimental results. Such a comparative study of both core modelling approaches for bending analyses has not been described in the literature before and can be considered as notable contribution to the
state of the art.
79
Figure 60 with the help of pictograms. The experimental results are characterized by little scatter in terms of bending stiffness (below 5%). The scatter in terms of strength is
slightly more pronounced (up to 10%). An exception is specimen B19-L101-03, which is
an outlier in terms of strength due to premature failure in form of local core indentation,
while the other two tested specimens of the same configuration indicate face sheet failure. The obtained test results generally reflect the expectations and thus appear to be
plausible. Thin panels with comparably weak face sheets (B10, B07) fail due to face sheet
rupture, while the B26 and B19 panels fail due to local core indentation, unless loading
plates were applied to prevent this. Core shear failure is only evident with strong face
sheets and in W-orientation where the core shear strength is the lowest.
5.1.2 Numerical analysis
Due to the importance of bending tests in sandwich construction, there are several numerical studies on flexural bending of honeycomb sandwich beams available in the literature. These studies can be subdivided according to the applied core modelling approach. Heimbs [Hei06] and Zinno et al. [Zin11] model flexural tests of Nomex honeycomb sandwich beams using 3D-continuum elements for the core. Their models are intended to enable validation of existing material properties, while the focus is on the
global panel response. Other models are targeted towards localized effects such as indentation or wrinkling. They therefore apply detailed meso-models. Giglio et al. [Gig12]
investigated indentation in case of three point bending of a Nomex core with aluminum
face sheets. They conclude that the friction parameter between load cylinder and face
sheets has significant impact on the indentation pattern of the simulation. They suggest
a friction parameter between 0 and 0.3. Staal [Sta06] studied face wrinkling of Nomex
honeycomb panels with GFRP face sheets using detailed meso models and 3D-continuum models. He reports that a meso-model has no benefits over 3D-continuum models
for the prediction of face wrinkling. In the present work all ten tested configurations are
implemented in a virtual testing framework. The objective is to validate and benchmark
previously derived modelling approaches and material properties. Therefore, the bending tests are implemented at meso-scale as well as using 3D-continuum elements for the
core. The experimentally observed failure modes enable to validate and benchmark the
model performance regarding the out of plane compressive and transverse shear
strength of the core models. In addition, the compressive strength of the face sheets is
calibrated and the implemented modelling approach for the face to core bond is evaluated. In the following, the implemented models are described in detail before the simulation results are presented in comparison to the experimental results. Such a comparative study of both core modelling approaches for bending analyses has not been described in the literature before and can be considered as notable contribution to the
state of the art.