Changing Beam Orientation - Multiple Support Structure

[JohnHerrman2382]

Attached is a "PDF" file depicting a support structure being evaluated.  It is a four leg (vertical beam) structure attached to a steel band.  The steel band and vertical support members were all generated in the Algor Simulation software.  To simulate the actual loading conditions, the W beam members at 0 and 180 degrees (top and bottom of the structure) need to be rotated 180 degrees.  When rotating these members using a delta move of the "K" node, I'm receiving multiple errors calculating the model.  I am assuming that it is a procedural issue having to do with the way I'm defining the rotation.

On a second issue is it possible to offset, the vertical beams to more closely represent the beam attachment to the flange edge rather than the beam centerline.

Thanks for the assistance.

[John_Holtz]

Hi John,

You can rotate the two vertical beams by selecting the lines ("Selection > Select > Lines") and changing the surface number to 3 ("Geometry > Tools > Modify Attributes"). The first few surface numbers orient the beams so that axis 2 is pointing in these global directions:

• surface 1, axis 2 in Global Y
• surface 2, axis 2 in Global Z
• surface 3, axis 2 in Global X

See the section "Beam Element Orientation" on this page in the Help: "Autodesk Algor Simulation > Setting Up and Performing the Analysis > Setting Up Part 1 > Linear > Element Types and Parameters > Beam Elements".

Beam elements can be offset by selecting the lines, right-click, and choosing "Add > Beam Offset".

[xli]

Hi, John,

Thanks for John Holtz has told you a way easily changing orientation of a group of beams by changing their lines surface color to realign orientations with correct default node. 

Here is another way, pick up the beam elements (lines) to define a new k-node instead of the default k-node, then right click mouse to choose "Beam orientations", then pick up "new" from sub-menu to create new k-node, input the coordinates of the k-node you want globally. Then return. You should see a circle representing the k-node and a green arrow pointing to it for the 2-direction of the beams.

I am also interested your way of "delta" defining k-node. Would you please to enlighten a little bit of it, especially the error messages you got? That may help us to improve, thanks!

-xli

[JohnHerrman2382]

I'm attaching a "PDF" file depicting the error messages as well as the "K" coordinates utilized in this model.  The procedure that I'm following to orient the "K" nodes on these two members seems consistent with that described above.  However, I'm receiving the message that the analysis did not finish - there were 0 warnings and 2 errors during this analysis.  If attempting to view the results after receiving the above message, I receive the second message (0x6B: Invalid element number 1 in part 6: unexpected 2-node degenerate configuration).

Thanks for your assistance.

[JohnHerrman2382]

These modifications worked and provided the visual model I was intending to evaluate.

For a better understanding of the results, can you advise how the beam offset is handled in the evaluation.  As I understand the explanation in the Help section, the analysis considers a solid plate joint between the connecting members.  Does this plate section add rigidity to that area of the juncture?  Does the connecting plate end at the bottom of the steel attachment ring, allowing the resulting offset moment to transfer into the lower leg section and be included in the evaluation?  I've attached a "PDF" snapshot from the results page to help clarify these questions.

Thanks again,

[xli · Accepted solution]

Thanks for providing your failure case for us to investigate. (I will dig into it and see what happened. I could not simply spot anything was wrong yet). 

For your followed question: you can think about beam offset is a "rigid" connection from the beam to the global structure. The connection is full, i.e. including all translational and rotational DOFs. If your real connection need to consider certain DOF is loosed, you may use beam endreleases there to achieve. Therefore, loads like beam distributed loads etc. are correctly applying to global structure though the connection and the deformation in whole structure reflects to the offset beams too by a "rigid" bar transformation.

-xli

[JohnHerrman2382]

Considering the application in question, where the top steel band is not supporting the leg structure but basically providing a means of load transfer to the legs, would you release the DOF for everything but the vertical translation?  Would you release all the DOF on the lines in the lower section of the leg members?  Would all the DOF changes made to I nodes be included on the J nodes?

Thanks again,

John

[S.LI]

1.) beam offset:

To apply beam offset will change the loads on legs. In your problem, without offset, the leg seems under a simple compression load to support the ring. With offset, loads becomes simple compression + bending, which easily make legs unstable (buckling).

2.) DOF release:

To release DOF or not corresponds to different work cases. No release is for bonded connection such as welding, and release is for pin joint or some other connections, which depends on how DOFs are released . According to your project, you can choose the right one. Or you can compare the stress results by simulating several of them.

[xli]

John,

You may set beam endrelease on four beam elements (each for a leg) at the end node connecting to the ring. Those could be I-node or J-node, you have to make sure, endrelease should be set on correct one. A closed circle on the element represents which got the your setting, i.e. telling which end is I-node and which end is J-node. Since you only consider the verticle loads is transfered vs the connection, you may leave only Tz DOF as un-released in the endrelease setting.

But with only vertical connection, that "ring" may not be well constrained so it might also need some extra fixities using weak boundary elements or symmetric BCs.

-xli

PS, I have take a second look on the PDF file you sent for the k-node setting problem you once had. But I could not see anything caused beam error you got. If you can post the model as it is (with the error), that would be great! Thanks.