asdeasoft.net

Hi,
Hope you are doing well.
I have a question about my STKO.
I am receiving the following errors when running the program but I don't know what is causing them.

(WARNING BandGenLinLapackSolver::solve() -factorization failed, matrix singular U(i,i) = 0, i= 0
WARNING BandGenLinLapackSolver::solve() -factorization failed, matrix singular U(i,i) = 0, i= 0
WARNING BandGenLinLapackSolver::solve() -factorization failed, matrix singular U(i,i) = 0, i= 0
WARNING BandGenLinLapackSolver::solve() -factorization failed, matrix singular U(i,i) = 0, i= 0
WARNING BandGenLinLapackSolver::solve() -factorization failed, matrix singular U(i,i) = 0, i= 0
WARNING: CTestNormDispIncr::test() - failed to converge
after: 1 iterations current Norm: inf (max: 0.01, Norm deltaR: nan)
WARNING: CTestNormDispIncr::test() - failed to converge
after: 34 iterations current Norm: inf (max: 0.01, Norm deltaR: nan)
NewtonLineSearch::solveCurrentStep() -the ConvergenceTest object failed in test()
StaticAnalysis::analyze() - the Algorithm failed at step: 0 with domain at load factor 2e-06
OpenSees > analyze failed, returned: -3 error flag
ERROR: the analysis did not converge
while executing
"error "ERROR: the analysis did not converge"")

I would like the STKO model to be like the tcl file attached in the rar file below. In the tcl file I do not receive such errors. Would you please help me with this?

Thank you,

Hi, It should work in both cases correctly. However, the error says there is some connection issue (leading to singularity matrix) between node or elements. I tried to assign elastic element (with large E and G values) instead of rigid links (Rigidlink_R and Rigidlink_L) in your model, the model ran without errors.

Would you send the file you said worked without errors? because I tried that but again I am receiving some errors.

Sure, As you'll see in the attached file, the 'rigidLink' option is creating an issue when the tendon and beam nodes coincide. I've made a temporary fix by adjusting the tendon's position. I'll be working on a more clean solution and will keep you updated.

Thank you so much.
How about the middle rigid link. Because there has to be a rigid link in the designated location. I saw your model but there was no rigid link in the middle.
In the OpenSees (TCL language) I used the same coordinates for the nodes creating the middle one but there was no issue.

Yes, I did not add the middle link at there was already fixity at the bottom and Uz zerolength element. Just remember, rigidLink-beam doesn't make sense if the two nodes are coinciding, you can use equalDOF if that is the case.