Hi everyone,
I have created a model of a 6-storey steel MRF with braces(BRB). My intention was to simulate the connection of each brace to the beam-column joint along with all other classic elements(gusset plates, etc). The braces are simulated as displacement based elements.
My question is whether the modelling approach, using a zerolength element with the appropriate orientation across the brace is correct,and by no means lesser than using a corotational truss element. The reason I am asking is because the Constraints I imposed to the zeroLength element (remember it is supposed to be an axial spring) is "equalDOF" in"-dir 6" and the element is free to deform in dirs 1 2 (because of the 45 degrees orientation, I need both vertical and horizontal components of displacement to be free=global U,V, so I get the resultant=local U). My skepticism is due to the fact that equalDOF refers to the GLOBAL coordinate system, and not to the local, that's why I released these dofs.
I don't use a corotational truss element, because it requires an area A, while I have a brace and my stifness is EA/L, but my material is steel02 & Hysteretic in parallel (Tri-linear plastic behaviour, different in tension and compression, with combined isotropic and kinematic hardening),so Hysteretic material messes stifness up.
I couldn't find something similar in the forum or anywhere else and that's why I'm asking. Most people use corotational truss for brace elements and other elements related to them. The results obtained so far, are controlled with similar models in other software and the difference is less than 5% for zeroLength, but very bad for corotTruss (I believe because of the material).
These models are being validated through monotonic pushover and will be subjected to IDA.
I would like your opinions on the matter. Also, is there a chance that corotTruss does not cooperate well with displacementBased beams (braces element) and that maybe is the reason for the bad results?
Any views on the subject are most welcomed!
The model is huge, and it would be pointless to post it here, so I will just post what is relevant to the axial spring approach(first storey):
***********************************************************************************************************************************************************************************************
...Code...
Material:
uniaxialMaterial Hysteretic 2 750 0.0055 856 0.0085 3731.37 0.3 -750 -0.0055 -856 -0.0085 -2605 -0.3 1 1 0 0 0;
45 degree element:
element zeroLength 1000 1 2 -mat 2 -dir 1 -orient 1 1 0 -1 1 0;
135 degree element:
element zeroLength 1006 20 2001 -mat 2 -dir 1 -orient -1 1 0 -1 -1 0;
...Code...
[END]
**********************************************************************************************************************************************************************************************
Inclined zeroLength element, on brace
Moderators: silvia, selimgunay, Moderators
Re: Inclined zeroLength element, on brace
when the truss is orientated, a transformation matrix is created. the truss optains the nodal displacements and then hits it with the transformation matrix to determine deformations in the different local directions.
the equualdDOF is just imposing the noal displacements. if 2d model, to constrain the rotations you would want dof number 3 and not 6 (even though the genius ehind the zerolength would require you to specify a 6 for this same dof).
the equualdDOF is just imposing the noal displacements. if 2d model, to constrain the rotations you would want dof number 3 and not 6 (even though the genius ehind the zerolength would require you to specify a 6 for this same dof).
Re: Inclined zeroLength element, on brace
I forgot to mention that I checked the braces internal forces and there was only axial (good). I also checked the global displacement (& forces) horizontal and vertical components and these agreed with the local expected ones (in the elastic region) = orientation of zerolength is good.
However I didn't understand the truss orientationconcept, I thought corotational truss is oriented by itself and doesn't need any input.
I understand the comment about equalDOF.
I see no comment about the zerolength approach, so I guess it is ok to proceed.
P.S. To imagine what I try to do in a better way, think of dampers=axial springs at the ends of the brace.Similar concept but with zerolength, free to deform in -dir 1 2.
However I didn't understand the truss orientationconcept, I thought corotational truss is oriented by itself and doesn't need any input.
I understand the comment about equalDOF.
I see no comment about the zerolength approach, so I guess it is ok to proceed.
P.S. To imagine what I try to do in a better way, think of dampers=axial springs at the ends of the brace.Similar concept but with zerolength, free to deform in -dir 1 2.