Convergence to unconvergence after replacing con06 by con01

macropaine · Post by **macropaine** » Mon Jul 27, 2009 6:55 pm

Hi, Dr. Silvia.
When I utilize the element dispBeamColumnInt with concret06, the anlaysis converges but the result is unreasonable. So I replace the concrete06 by concrete01,
...
#uniaxialMaterial Concrete06 $confconcr $fc $eo $n $k $alphaC $fcr $ecr $b $alphaT;
uniaxialMaterial Concrete01 $confconcr $fc $eo $fc [expr 1.6*$eo]
...
system BandGeneral;constraints Transformation;numberer RCM;test NormDispIncr 1.0e-12 10 3;algorithm Newton;
integrator DisplacementControl $lpoint 1 $dU 1 $dU $dU

the error info is:
WARNING BandGenLinLapackSolver::solve() -LAPACK routine returned 2
WARNING NewtonRaphson::solveCurrentStep() -the LinearSysOfEqn failed in solve()
StaticAnalysis::analyze() - the Algorithm failed at iteration: 0 with domain at
load factor -1.#IND

If I use ModifiedNewton algorithm:
system BandGeneral; constraints Transformation; numberer Plain;
test NormDispIncr $tolLatNew $iterLatNew;
algorithm ModifiedNewton -initial

error info is:
WARNING BandGenLinLapackSolver::solve() -LAPACK routine returned 2
DisplacementControl::newStep(void) - failed in solver
StaticAnalysis::analyze() - the Integrator failed at iteration: 0 with domain at
load factor 0

Could you please help me to solve the problem?
Thank you very much!

silvia · Post by **silvia** » Mon Jul 27, 2009 7:01 pm

try concrete02 and let me know.

macropaine · Post by **macropaine** » Mon Jul 27, 2009 7:46 pm

Thank you for your fast reply, Dr. Silvia.
If I use concrete02, analysis converges. However, the analytical value of load is unreasonably little. The aim for using concrete01 is to decrease the slope of declining branch of the load-displacement curve after crack or yield. Because using concrete06, the decling branch is very steep.
Could you please give me advice to make the analysis to be convergence?
Should I change the constrain methods or system methods?

silvia · Post by **silvia** » Mon Jul 27, 2009 9:19 pm

it all depends on your model, yes. a steep change of stiffness can cause significant problems.
have you looked at the scripts in the examples manual? there are scripts on how to deal with non-convergence, you make some changes to the analysis.

macropaine · Post by **macropaine** » Tue Jul 28, 2009 8:13 am

yes, I have employed the script in the examples manual for non-convergence.
I think I have to change the shape parameter n of concrete06.
Thank you, Silvia.