Discussion Groups

Simulation Mechanical and Multiphysics

Reply
joe
Member
joe
Posts: 4
Registered: ‎03-30-2011

Results When Using Beam End Releases

1117 Views, 8 Replies
09-04-2012 05:06 AM

I am modeling a steel structure (a framework of W-beams) using beam elements in order to perform a vibration analysis (modal and frequency response).  The connections between the structural steel have been designed such that they are not moment connections, but rather have been braced as needed (knee braces etc.).  In order to model this accurately, I have used beams with end releases at their intersection (connection).  Using guidance from help files and tutorials, the beams are also split into multiple segments in order to get more nodes and more accurate results.

 

When we apply a rotational end release (release all rotational DOF, but keep the translation DOF) the beam segment that gets released always has a very large moment (much larger than the rest of the structure).  The rest of the beam (ie. the rest of the segments that the beam is composed of) behaves as it should - roughly as a simply supported beam.

 

I have tried breaking the beam up into much smaller pieces, and I continue to get the high moment results in the first segment.  I have also tried changing the End Release options, and removing them completely to troubleshoot.  When the End Releases are removed (or disabled), all of the beam segments behave correctly, including the first segment with the fixed connection.

 

When the end of the beam gets released to rotation, I don't think that there should be a moment present in the beam segment at the released end - what am I missing?

 

I am currently using AutoDesk Algor Simulation 2012 and 2011, and have included a simple test file that was made to test the situation.

Please use plain text.
Employee
S.LI
Posts: 398
Registered: ‎05-06-2010

Re: Results When Using Beam End Releases

09-05-2012 08:47 AM in reply to: joe

I think your result is all right.

One of the beam end releases (BER) is applied between element 1 and 2. Another one is between element 3 and 4.

it means the segment including elements 2,5,6,7,8,3 is a simple-supported beam now.

Under distributed loads, non-zero bending moment is all right here.

 

Also, you'll see zero local 3 moment at the joint node between element 1 and 2, which means "release".

----------------------------------------------------------------------------------------------
If this response answers your concern, please mark it as "solved".
Please use plain text.
*Expert Elite*
AstroJohnPE
Posts: 576
Registered: ‎08-30-2012

Re: Results When Using Beam End Releases

09-05-2012 11:45 AM in reply to: S.LI

I agree with S.Li that the moment diagram for a static analysis looks like I would expect. (See attached.) Perhaps we are not understanding your written description.

 

For a fixed-fixed beam with a distributed load under static stress, the moment diagram looks very similar to what I have shown. That is, there are two locations along the length where the bending moment is zero. If you make a hinge joint (end release) at those two locations, the moment diagram does not change at all! Whether a coincidence or not, that is what you have in your design.

 

Of course, the modal analysis will be different between a beam without end releases and a beam with end releases.

 

John Holtz, PE
Mechanical Engineer
Pittsburgh, PA

Simulation Mechanical user since Dec 1997
Please use plain text.
Member
TimDewhurst3723
Posts: 3
Registered: ‎04-16-2013

Re: Results When Using Beam End Releases

04-16-2013 02:41 PM in reply to: AstroJohnPE

The original problem description appears to be exactly the problem I am having.  However, when I open the file and run the analysis, the bending moment diagrams seem correct.

 

Attached is another case where I have used beam end releases at the top of two arms.  For the elements where the end release is applied, the bending moment is huge.  This is true whether the arm consists of 10 elements, or a single element.

 

If there is an end release at the end of the element, then the bending moment should go to zero as there is nothing to constrain the rotation.

 

I would appreciate feedback on this model.

 

Thank you.

Please use plain text.
*Expert Elite*
AstroJohnPE
Posts: 576
Registered: ‎08-30-2012

Re: Results When Using Beam End Releases

04-17-2013 06:35 AM in reply to: TimDewhurst3723

Hi Tim,

 

The results are correct in version 2013 Service Pack 2 (SP2), and presumably with the current version 2014. I suggest that you update your software from 2013 no SP. If I recall correctly, there were a number of solver bugs fixed in the first and/or second service pack. (The "read me" files on the web page for each service pack describes what changes are in each.)

 

John Holtz, PE
Mechanical Engineer
Pittsburgh, PA

Simulation Mechanical user since Dec 1997
Please use plain text.
Distinguished Contributor
m.granata
Posts: 154
Registered: ‎04-07-2005

Re: Results When Using Beam End Releases

04-27-2013 11:27 AM in reply to: joe

I am still use pre-Autodesk, Algor software user, but it sounds like you are seekings results that a truss member and not a beam member would provide.  Basically just a member that can support axial compression and tension loads.  Truss elements may be suitable when you are seeking only pin type conditions at the nodes.  That's the approach I have used with both Algor and Ansys FEA software.  Then verify it with a few simple stress/deformation calcs.  Good Luck.

Please use plain text.
Member
TimDewhurst3723
Posts: 3
Registered: ‎04-16-2013

Re: Results When Using Beam End Releases

04-27-2013 12:20 PM in reply to: m.granata
The problem appears to be internal to the software. Updating to 2013 SP2
eliminated the problem with the beam end release.

While truss elements would be appropriate for the 2-force members, beam
elements are necessary for the multi-force arms.

Thank you for your input.

--
Dr. Tim Dewhurst, P.E.
Professor of Mechanical Engineering
Solar Boat Team Faculty Adviser
Cedarville University
251 N. Main Street
Cedarville, OH 45314
(937) 766-7654
dewhurst@cedarville.edu
Please use plain text.
Member
TimDewhurst3723
Posts: 3
Registered: ‎04-16-2013

Re: Results When Using Beam End Releases

04-27-2013 12:21 PM in reply to: AstroJohnPE
Thank you very much for your response.

We upgraded to 2013 SP2 and the problems have disappeared.

Thank you very much for your response.

--
Dr. Tim Dewhurst, P.E.
Professor of Mechanical Engineering
Solar Boat Team Faculty Adviser
Cedarville University
251 N. Main Street
Cedarville, OH 45314
(937) 766-7654
dewhurst@cedarville.edu
Please use plain text.
Distinguished Contributor
m.granata
Posts: 154
Registered: ‎04-07-2005

Re: Results When Using Beam End Releases

04-27-2013 01:08 PM in reply to: TimDewhurst3723

Sorry, I guess I misunderstood your model.

Please use plain text.