Problem dealing with double angle section

[gortiz5PR4T]

Hey everyone, I'm here again asking for help.

I'm designing a frame so I need to implement a section for some members with doubled angle cross section, that would not be a problem, but I need to be able to design and verify it with the steel module, and the thing is that I'm using AISC, so no considerations are available for that type of sections which are reviewed on that design code. And make doubled angle sections in section definition module is not an option, since I'm working with a lot of different angles.

The approach I took consists of making a "second frame" 1 inch separated, and then make a rigid link between nodes, but I cannot make it works, while verifying displacements, second frame seems to be completely free, so displacements are so big.

Some hints, members (except columns) have truss property defined, cuz it's needed to work only for axial load, and slope and ground level of columns was part of project requirements. If needed, model was attached to the topic.

Do you guys, have an idea of what I'm taking wrong?

[Rafacascudo]

" Truss " bars cannot be attached to another "truss" bar in a 2d frame , 3d frame ou shell structure type.

On a standard truss structure at least the chords have to be regular bars with no releases. Then the diagonals can be assigned as "truss" bars

That is probably why you are getting instability type 1 warning , which cannot be ignored

For the double angles steel design issue , check here

https://forums.autodesk.com/t5/robot-structural-analysis-forum/steel-design-compoud-sections/m-p/435...

[gortiz5PR4T]

Thank you for your reply @Rafacascudo , I tried the way you mentioned, but even with that, I feel like there's still something I miss in the way, due to the extremely exaggerated displacements I'm getting. I think, the main problem is the "attaching" of the substructure is not done correctly (i'm using rigid links in the nodes the frames join with the columns), I tried many ways, but none worked. It's supposed I should only get 10 and 4cm of x, z, displacements (checked with compound sections).

Making it well, it's supposed that the thing of "design" a member with sections of doubled angles back-to-back with a span other than the pre-loaded sections in code's database, wouldn't be required.

[Rafacascudo]

You should follow the advices of Artur Kosakowsky in the link I sent on my previous post especially here and model it as a single section

"The sections from the databases are assumed as solids (including the double angles with spacings) whereas the compound section definition is primarily intended for sections that are build from of more than one separate 'parts connected with battens or diagonals certain points along bar length. They can be code checked (including specific provisions for such section types) with selected steel design codes (e.g. EC3).

If you define the double L 3x2.5x0.25 with no spacing (solid) using the compound section definition and pressing the weld (triangle) symbol next to the spacing field the section will be checked. Assuming that the double L sections in your model are under axial force and unidirectional bending in the Z direction you can redefine them like that which should have no influence their code verification."

Send your updated model so I can have a look

[Rafacascudo]

I deleted the "2nd truss" and kept the existing structure of the main axes plane.

Also deleted all rigid links and redone the double angles as compound sections as Artur Kosakowsky advised on the link I sent previously.

Model ran with no warnings and errors .

  Then I proceeded to the AISD LRFD 2000 steel design. Assigned some fictional code parameters to diagonals and chords , which you will have to correct accordingly to the real structure you have .

Double angles were checked by the steel design module(AISC LRFD2000)as Artur predicted

Test model attached . As I said before ,it is just a test model to see if the double angles could be checked as Artur advised .You still have a lot of work to do concerning the steel design( code parameters for chords and columms , loads , calculation configuration , etc))

[gortiz5PR4T]

Thank you for your help @Rafacascudo , and sorry for the late answer. The thing is that I need to "simulate" a real behavior of the bar with double back-to-back angled section with the span of 1in, I understand that the program can run the verification for non-spaced angles, but that span helps a bit with Iz of the cross section.

Also, I know the difference is too small, but do you know why the strength of the cross section is better if you check it as a compound section and with for example, section in AISC database of two angles back-to-back? I bet the calculated properties of the solid section are more precise in one scenario than in the other.

[Rafacascudo]

Really neglegible ratio difference.  Lay are identical ,so maybe the bending moments are a litle different.

Check the detailed results to see where the difference is