equalDOF in 3D models

[fh]

Hi
I wanted to ask about any probable problem using equalDOF for modeling pinned connection (beams or braces) in 3D modeling. I asked it because I was trying to run a simple 3D modeling. I was successful when the connections were all rigid, but while adding braces and using "equalDOF 1 21 1 2 3'' (for example) I got this message:

WARNING BandGenLinLapackSolver::solve() -LAPACK routine returned 26
WARNING NewtonRaphson::solveCurrentStep() -the LinearSysOfEqn failed in solve()
DirectIntegrationAnalysis::analyze() - the Algorithm failed at time 0.02
OpenSees > analyze failed, returned: -3 error flag
-3

I will be thankful if you can help me.

[vesna]

equalDOF command is the right one for defining pins. You probably made a mistake while defining pins, so the solver can not find a solution.

I would suggest you to have torsional degree of freedom constrained along with the translations for each pin.

[fh]

Dear Vesna;
In another approach I tried using zerolengh element for rotational DOFs. But now I have another problem analyzing my 3D model. I got the message:
''Ran out of memory for A size''
I wanted to ask if there is any limit for the number of elements or others in Opensees. It should be said that I have a 10 story building with 5 spans in each direction which 3 of them are braced.

Thank you in advance.

[vesna]

Make sure that if you are using fiber elements you do not have too many fibers. Also use sparse solver to solve system of equations (UmfPack or SparseSYM).

[alirezaii]

Hello vesna and fh
I have a same problem with equalDOF in 3D models. I develop tow nodes with identical cordinate and use "equalDOF 313 31301 1 2 3 4 5 6" for these, but OpenSees give me different displacement in output along with many warning during run.

[pp_0928]

I have a same problem with modeling pins in two horizontal direction in 3D moment frame. If I use equalDOF to connect pins, two nodes are isolated. If I use zerolength element, it ran out of memory. I already tried UmfPack and SparesSYM. They don't work neither.

Hope that we can have a positive solution in this post. Thank you guys.