shell from ribbed deck3.gh (66.8 KB)
Hi everyone,
I’m trying to do some FEA analysis in Karamba for timber stressed skin panels that have two plywood panels and lumber ribs in between. I’m not able to make a slab that has a composite actions and it seems like in my model only ribs are in action as beams and the shells (plywood panels) are not involved in bearing any stress.
Has anyone modeled composite cassettes before? or can anyone help me with this?
Best,
Nasi
Dear @n112,
your definition does not work because the beams are not connected to the upper and lower decks. For elements to be connected in Karamba3D they need to share nodes. For embedding beams in shells you could use the inclusion points input (IPts) of the Karamba3D MeshBresps-component and supply the endpoints of the ribs. The latter should be subdivided into small pieces so that they are sufficiently connected with the plate.
A beam-eccentricity allows to offset the beam axes from the center of the plate. If you connect the upper plate to the ribs the lower plate could be simulated by having I-cross sections for the ribs where the upper flange width is zero and the lower flange width equal to the contributing with of the lower plate.
Alternatively the whole system could be simulated using shell elements only. In that case make sure to have a sufficient number of shells along the depth of the ribs. Otherwise the system will be too stiff.
– Clemens
Thank you so much for the feedback. I tried extending the lower flange of the I-beams to create the lower panel, but still its seems not involved in analysis or too stiff.
I also couldn’t perfectly align the beams upper end with the panel’s centerline (screenshot attached), which might be the reason that system is not working well.
shell monday.4.gh (291.4 KB)
I’ve attached the new GH file.
Dear @n112,
here is an example how to attach a rib to a plate:
Plate_WithRib.gh (42.6 KB). I hope this helps.
Be aware of the fact that the effective width of the lower flange is limited: Due to in-plane shear deformation the normal stresses are not uniform and decrease with the distance from the web.
– Clemens
