NonLinear axial test with Zerolength Element didn't converge

[Jean Domínguez]

Please FMK, tell me what i´m doing wrong in this test. It´s a simple column that is loaded to 2500 Kgf with a linear analysis and next is extended with displacement control (Pushover) to a displacement a bit bigger than the supposed yield displacement. The algorithm stop when the model is near the yield point. here is the script.
/code
# Units Kgf-cm
wipe;
model basic -ndm 2 -ndf 3;
file mkdir colt;

node 1 0 0
node 2 0 0
node 3 0 300

fix 1 1 1 1
fix 2 1 0 1

uniaxialMaterial ElasticPP 1 285600000 0.00117
geomTransf Linear 1;
element zeroLength 1 1 2 -mat 1 -dir 2
element elasticBeamColumn 2 2 3 1200 238000 160000 1;

recorder Node -file colt/DFree.out -node 2 3 -dof 1 2 3 disp;
recorder Node -file colt/RBase.out -node 1 -dof 1 2 3 reaction;

pattern Plain 100 Linear {;
load 3 0.0 2500 0.0 ;
}
constraints Plain;
numberer Plain;
system BandGeneral;
test NormDispIncr 1.0e-8 6 ;
algorithm Newton;
integrator LoadControl 0.1;
analysis Static
analyze 10;
print colt/colt.out

loadConst -time 0.0

puts "Model Built"

pattern Plain 200 Linear {;
load 3 0.0 1.0 0.0 ;
}

constraints Plain;
numberer Plain
system BandGeneral
test NormUnbalance 1.e-8 6 0;
algorithm Newton;
integrator DisplacementControl 3 2 0.01
analysis Static

set Nsteps [expr int(0.40/0.01)];
set ok [analyze $Nsteps];
puts "DonePushover"
/code

[fmk]

of course it's going to stop converging at the yield point .. you are using elasticPP which has a zero tangent after yield, thus your tangent matrix will be singular. so either switch to ModifiedNewton with initial stiffness iterations or change your material to have a small positive (or neg slope assuming you choose an appropriate solver).

[Jean Domínguez]

Thanks fmk, obviously EA/L=0 at yield point thanks, seems to be very difficult to see simple things when one is concerned about learning another, i've used the hardening material and was flawless.
Another question. Have you (or anyone reading this post) a working script of "clough" peak oriented uniaxial material ?, i´m trying to get the histeresys graph but the model never do nonlinear excursions. i readed the Altoontash Dissertation, and he mentioned 16 parameters but don´t show neither any example nor detailed explanation of each parameter. ThX.

[fmk]

i have asked Arash to respond.

[fmk]

arash's response:

For Peak-Oriented Clough model, eight parameters define the
force-deformation backbone (see figure 3-11, Altoontash thesis) and eight
parameters define the strength deterioration, stiffness deterioration,
accelerated stiffness deterioration and cap deterioration. for more
information regarding the Clough model please refer to following references:

Ibarra, L., Medina, R., and Krawinkler, H. (2004). “Hysteretic models that
incorporate cyclic strength deterioration and stiffness degradation”,
*Earthquake
Engineering and Structural Dynamics*.

Ibarra, L. (2003) “Global collapse of frame structures under seismic
excitation”. Ph.D. Dissertation, Stanford University, Department of Civil
and Environmental Engineering.