Element Orientation Vector 2

[kacstic]

I just started using NEiNastran v9.02 recently and for practice, i am modeling basic line models (steel beam structure for example).

When I mesh each line (or curve), I designate the material, beam cross section, and then it asks for the element orientation vector. I presume this is to identify the major and minor axis of the cross section.

However, when I have a beam that is not horizontal or vertical, obtaining the correct element orientation vector is perplexing. Is this step crucial in calculating the global stiffness of my structure? For now, I designate a unit vector in the x, y, or z axis that is not collinear to the beam's length. Thanks for the help...

 

[SWComposites]

Your presumption is correct. The orientation vector is critical unless the beam cross-section properties do not vary with orientation, e.g., a circular section. Have you read the user's manual re beam element definition?

 

[kacstic]

I have read that part but it is no help, unless I am reading the wrong one. Do you know of a better reference source than the one NEiNastran supplied? Thanks!

 

[SWComposites]

The figures in the CBEAM and PBEAM sections of the NEi Reference Manual are fairly clear. See also the Bar Element section in the NEiModeler (Femap) Users Guide.

 

[GregLocock]

Try switching on the beam orientation vector in the display menu of your preprocessor. It'll make things slightly more clear.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[kacstic]

Thanks! Those references helped my understanding a little more. Also the visualization of the orientation vector was helpful.

 

[GBor]

kacstic,

How about giving them a "purple star". In the lower left hand corner of all posts is a purple star and a "Thank ______ for this valuable post". Click on that for each person that gave you a valuable post and they get a vote for "Tipmaster of the week".

Thanks!

 

[RobynC]

Hi I have a similar question to the one discussed here.

I am trying to mesh a line (set of curves) that uses all three golbal planes with beam elements. Is there a way I can make the element orientation vector to be always perpendicular to the direction of the curve, rather than being relative to the global coordinate system?

At the moment I get errors as at some points since the "element orientation is colinear with element x-axis". I do not want to define a random out of plane vector as the element orientation is important to the analysis.

 

[GregLocock]

I think you need to tell us which preprocessor you are using. At least in God's preprocessor this is fairly easy, Hypermesh, not so much. Clue: a vector can be very long.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[RobynC]

Oh sorry, yes that would help. Am using FEMAP

 

[GregLocock]

Good choice - So just set a node a long way from your structure and use that as your reference vector

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[RobynC]

Thanks, will give that a go