topic Re: want to simulate slab plane rigidity in Robot Structural Analysis Forum

want to simulate slab plane rigidity

achillesgr — Sat, 06 Jun 2015 19:42:58 GMT

Help, I have the typical metal structure with beams and purlins that supports a wood floor .I simulate it with a cladding upon purlins.the robot does not recognize a plane rigidity and it gives me large horizontal deformations at the middle of the beams. How can I tell the robot that I have that kind of plane rigidity? Is the cladding good choice for floors or it is only for claddings? Also I did not see anywhere any end type supports for this "cladding" that sitts upon the purlins.is it necessary to give any end conditions and how? If I have a cantilever ,I give only one cladding or slab or two on each side of the beam? Thanks I appreciate any Hall.

Re: want to simulate slab plane rigidity

Rafacascudo — Sun, 07 Jun 2015 00:50:43 GMT

Claddings are objects only for load distribution.they don't contribute to the structure stiffness.
If you want to model the slab horizontal stiffness , you can create a panel instead ,and assign it a thickness.
This panel can work like a shell, or a rigid ,or flexible diafragm.you choose it.
Claddings are always "pinned".they use the trapezoid method to distribute the load.
Don't need several claddings. One giant cladding will do the job.Only pre-requisite is that it has to be totally flat

Re: want to simulate slab plane rigidity

achillesgr — Sun, 07 Jun 2015 11:07:30 GMT

Helo, Can I add a panel with the mensioned properties without deleting the already inserted cladding? I already assigned loads and generate combinations to the cladding and I wonder what happens to the combinations if I add a new panel now with new loads. Should I delete the combinations and generate them again? Can the cladding and panel exists together? Cladding for loading the purlins and panel for rigidity? Thanks!

Re: want to simulate slab plane rigidity

Rafacascudo — Sun, 07 Jun 2015 15:21:28 GMT

Just select the cladding and goto menu results/object properties. The contour dialog window will popup. In the botton right area change from cladding to panel. Then just assign the correct thickness and panel model you want. It will keep your original cladding number , so you can keep your load cases and combinations intact,
don't create claddings and panel one over the other

Re: want to simulate slab plane rigidity

achillesgr — Sun, 07 Jun 2015 21:00:42 GMT

helo,

please see the attached image and file.

i tried to recreate a new clading and also a new panel. both created with the four corners of the outside beams.the clading distributes fine the load to purlins but panel dont. it does not seems right to break the big panel at smaller ones in each purlin. what am i mising?

thanks again!

Re: want to simulate slab plane rigidity

Rafacascudo — Mon, 08 Jun 2015 11:50:56 GMT

" Claddings are always "pinned" "

Not panels . If you use panels and want them to distribute load like a cladding you can do it using adjusting settings on " panel calculation model .

But if the span slabs are not continuous , of course, you´ll have to divide the " giant slab" into smaller ones and apply linear releases to one common edge.

Re: want to simulate slab plane rigidity

achillesgr — Mon, 08 Jun 2015 15:45:06 GMT

helo,

tried to make some panel calculation models acording to your pic but here is what i found:

the only way to force it to do load ditribution (one or two way according to the last checkbox) is to choose NO FINITE ELEMENT type at the top.

if i choose FINITE ELEMENt no distribution is made.

i ended using my pics model for the flooring.please corect me if i am wrong.

how do you approach situations like that? is it nesessary to use finite elements? are the self weights and other properties of the given slabs or shels taken into acount?

in a floor like that , do you create individuals shels betwen purlins or what?

in the case you mension , can you please help with how to apply linear releases to common edges?

Re: want to simulate slab plane rigidity

Rafacascudo — Mon, 08 Jun 2015 17:50:17 GMT

1st of all , your mesh size is too big. set it to 0,50m. Mesh the panel and freeze.

yes , it is possible to have finite elements and simplified transfer of loads. If you have temperature loads applied to the slab , finite elements are mandatory.

In the same picture(taken from your model) , there´s the place to create linear releases ,if you need them.

Re: want to simulate slab plane rigidity

achillesgr — Tue, 09 Jun 2015 06:46:44 GMT

yes, the load distribution shows fine with the collor diagram BUT it seems that the perlins eventualy does not take any load.

here is one purlin without finite element option:

the 59knm seems right since there is a 10KN surface load at the mesh and the purlins distance is 2m .span is 5m so MY=20*5*5/8=62.5 (the diference from 59 i assume it is from offsets?)

here is the same purlin with the same situation and load if i choose the finite elements at the panel calc model:

FZ: ??

Also , here is the complete model without and with mesh option at the panel model:

Please note the BIG diference in MY 's of the main frame.

why that?

shouldnt the MY with or without the panels finite element at the main frame will be the same?

both senarios have the partial stifening XY. only change at the model is the Finite - No finite

If the purlin eventualy does not have distributed load from the panel, where is the small 3.22 MY comes from?

is it from the deformation of the panel that it is suported at the perimeter main frame (not purlins) and eventualy "sitts" upon the purlins?

thanks!!

Re: want to simulate slab plane rigidity

achillesgr — Tue, 09 Jun 2015 07:08:41 GMT

Strange, here is the "automatic generated forces" on the purlin WITH the finite elements witch is OK , and the MY is 3.22?????

what is going on??

thanks

Re: want to simulate slab plane rigidity

Rafacascudo — Tue, 09 Jun 2015 12:52:30 GMT

Don´t forget the moment resisted by the slab if you choose finite elements

If you make a 2meters panel cut at the middle of the span you´ll get an integral moment of 47.57 kN.m

add the beam moment and the effect of the beam offsets and you´ll have aproximately your hand made / no finite elements moment valor

serrilated Fz and My beam diagrams are typical in models with beam offsets