Hi all,
I am a student currently doing a project using grasshopper, I am having problems with mesh loads not transferring correctly from shell elements to beam elements. In the simplified model I have attached there is a shell element that is supported by two beams, when I apply a mesh load to the shell the load seems to get transferred to the beams and then into the supports, however when I add a thickness to the shell element the load seems to completely ignore the beams and just transfer straight into the supports through just the slab. I was wondering why this issue is occuring and how can is be fixed so that the load is transfered through both slab and beam.
Any help is appreciated as I am looking to optimise the depth of both the slab and the beams.
Thanks.
Load Transfer Problem.gh (48.0 KB)
Hi,
when I open your script it seems to apply the loads properly. What do you mean by adding a thickness to the shell. The Shell will always have a default thickness of 1cm if you do not define a cross section.
Apply Meshload onto beams:
Apply Meshload onto shell:
Hi Matthew thanks for the reply,
I will try to explain what I am looking for a bit clearer, I am looking to apply the mesh load to the shell element as if it were a floor slab, I then would expect the load to be transferred from the shell to the beams and then from the beams into the supports (As if the shell was resting on the beams and the beams are resting on supports).
From the image you shared where you applied the mesh load to the shell element it looks as though the forces are getting transferred straight to the supports through the shell and ignoring the beams that it is supposed to be resting on.
When I remove the cross-section of the shell and let it be its default value of 1cm it seems to do what I want it to do and transfer loads to the beams
Utilisation of shell with a cross-section defined:
Utilisation of shell without a cross-section defined:
Hi @16156374,
in your model the beams on the sides do take up loads from the shell:
Load Transfer Problem_cp.gh (50.8 KB) - otherwise there would be no shear forces in them. Since the plate is much stiffer than the beams they take more of the load.
– Clemens
Hi Clemens,
If the shell is sitting on top of the beam and mesh load is applied to the shell, should the load not be transferred to the beam as though it were a UDL along the beam? If this were the case the max shear force in the beam should be the same as the reactions at the supports which is not the case in my model.
I am also curious as to why the beams do not seem to be optimising as a UB 127x76x13 seems vey unrealistic for a 10m span with a concrete shell resting on top.
Thanks for the help, Oisin.
Hi @16156374,
the beams are connected to the plate at their endpoints. So the beam- and shell-elements act in parallel and carry external loads proportionately to their relative stiffness regarding each other. For getting out the reaction force at the support one therefore needs to add the shear forces in the beam- and shell-elements.
The reason for the small beam sections is, that the plate - which is assumed to be isotropic and linear elastic - takes the mayor part of the external loads.
You can try two things:
- include the plate thickness in the optimization.
- start the optimization with large initial cross sections for the beams so that they attract more forces at the beginning.
– Clemens