Separate loads with claddings or separate it as "planar on contour"

[Mirko.Jurcevic]

Hi,

I am doing some comparison between 4 cases of load distribution.

My small example model has two beams which have some loads distributed on them.

- In first example (leftmost) I've put a one-way cladding with 1,0kN/m2 on them and I've got what I excepted. Equal loads on both beams.

- Same result I got on the (rightmost) model - where I have two cladding connected in the middle of a span

- topmost example uses one cladding with two "planar on contour" loads and I also get expected results

Example in the middle is the one that I am concerned about:

- I've seen on a few places in the forum that people uses separated claddings for easier definition of loads (ex. wind loads on roof).

- A I can see, two claddings which are not connected in the middle doesn't distribute loads as I would expect (I want it to distribute in the same way as "planar on contour" example does). What it does, it distribute one cladding on one beam, and other cladding on the other beam. I am not sure if it should behave like that?

Model is in attach.

[t.sautierr]

after some test, I think it is due to the method for distributing the loads : it acts as cantilever when you divide you cladding (1 side bar -> cantilever and then one area of loading, 2 sides bar -> 2 areas of loading), no matter it is divided at the middle, at 1/3 etc ....... see second example bottom.

in the other hand, I think this is the purpose of countour loads ...

actually at no step one can specify the sides of a cladding with no support under are "connected" to make one big area, so not possible to have the desired effect ... except you overlap the pannel to obtain the desired value ... but useless to my opinion and false in terms of loading areas see frist example bottom 

what you can do is to use this (but this for simple loading and modification not easy like for contour ....)

[Mirko.Jurcevic]

Thanks for further analysis but I thought that dividing the claddings would make my life easier, what I get is quite opposite.

My idea (for further Robot development of cladding elements if there is no solution already) is that they have few more ways of load distribution.

Example (imagine those arrows that are showing load distribution):

this: ------>  or this: <-----  means it distributes loads only in one way along X (like cantilever - now they act like that) 

<-----> this means it distributes loads in two ways along X (like simple beam) - separate claddings connect to each other

With such diferent behaviours we can get separate claddings act like one (with this method <-------->) or every cladding for themselfs (with this ------>  or <------ behaviour) - like they do now.

[t.sautierr]

Yes or to be able to select one, two or X pannels to be linked and then RSA will apply the distribution load method on the "unified" area.

Bog!