buckling 2d autovalue problem via "custom" section

[GunMaker]

hi to all, I carried out a buckling analysis on a cross beam and I found that, during post processing, despite there were no errors or warnings during the analysis, almost all the autovalues that are derived from the "results summary" are equal. .
the way I made the fem model is as follows:

1)I created two keypoints at 2000 distance,after I joined them with a line.
2)I recalled the beam element 188 (cubic form).
3)I have created in separate seat a cross section of thickness 2 mm, height 70 mm and width 160 mm.
I then split the various lines that describe the perimeter of the section (as is written in the Help).
Finally I used the command "write from areas (secwrite) " and I selected the area.
4) I imported into the model beams the area via "read sect mesh"
5)I meshato the line created at point 1
6) I checked that the section was correctly imported via the "shape"
7) at this point I have bound the structure (simulating a cart to the loaded ester and a zipper to the unloaded end). Then I loaded the structure with an axial force of 15000 n in correspondence of the cart (on the keypoint).

8)I did first the static analysis (by setting the "calculate prestress effect") after which, once the static analysis has ended, I re-launched a new analysis "eigen buckling" setting in "analysis options" to extract the first 10 autovalues.

the exercise in question is as follows (if it were not clear to describe the problem):
the answer that gave me the program is that of the first image... obviously there is something that doesn't fit... but I just can't understand what I'm wrong. Did anyone ever encounter similar problems? How did you solve them? :frown:

thanks in advance for answers and availability :redface: .

 

[volaff]

I suggest you compare with your classmates and take a look at the exercises performed.

have you verified the coincidence of the knots of the mesh on the two orthogonal parts of the cross?

 

[GunMaker]

in fact the structure I have already solved it through another method (using shell elements, through extrusion)...but this seemed a valid alternative to test (in the future it could be less frustrating and convenient to use beam elements in place of continuous extrusions of geometries always equal)...any way how do I verify the coincidence of the mesh nodes (is there a method that prevents from the visualization of the structure)

 

[volaff]

Sure.
simply select the "commun" areas, take the elements in question then select the nodes and list the latter, verifying that there are no "double" nodes, ie with different numbering but with same coordinates: ansys evaluates them and manages them separately.

 

[GunMaker]

thank you volaff... I think I have identified the problem.. .evidently it's something I didn't understand in creating custom section....but I don't understand exactly what it is ... the Help suggests to create the section area, save it, set the size of the section lines and make the "write section from area" ...after creating the file that they went to open through the "read" command in "section/beams/custom sections" in the file of the structure... .

 

[mbindustry]

thank you volaff... I think I have identified the problem.. .evidently it's something I didn't understand in creating custom section....but I don't understand exactly what it is ... the Help suggests to create the section area, save it, set the size of the section lines and make the "write section from area" ...after creating the file that they went to open through the "read" command in "section/beams/custom sections" in the file of the structure... .

Have you solved it? I can't define the section with custom sections (as an element I have beam 188). asks me the value of iw (torsional hardness I think) but I don't know if it's the right way. how did you continue?