trilinear material behaviour

[Horr]

hi,

How can I define the moment-rotation behaviour of a rotational spring as Trilinear (no damage) or define with a polyline? Which material behaviour do you suggest?

Thank you

[vesna]

Hysteretic material looks like a solution for you:
http://opensees.berkeley.edu/wiki/index ... c_Material

[Horr]

Dear Vesna,

Thanks for your reply. But I need a material like steel01 (without pinching or damage). Is it possible by hesteretic material. I couldn't match the parameters to gain correct stress-strain relation.

[vesna]

If you set up parameters like this:

set pinchX 1;

set pinchY 1;

set damage1 0.0

set damage2 0.0

set beta 0.0

You will not account for damage and pinching

[Horr]

But I had tested these parameters. Now, I 've tested again. Here is the code and the result. It 's not what I need.

I wish I could send a stress-strain diagram (JPG file). Is there anyway to attach a file? please run and look at the result. There is stiffness degradation after load vector shifting (stress becomes negative or positive).

Regards

# --------------------------------------------------------------------------------------------------
# MATERIAL BEHAVIOUR
# HORR KHOSRAVI, 2011
#

wipe; # clear memory of all past model definitions
wipeAnalysis;
model BasicBuilder -ndm 1 -ndf 1; # Define the model builder, ndm=#dimension, ndf=#dofs
set dataDir Data; # set up name of data directory -- simple
file mkdir $dataDir; # create data directory

#################################### MOMENT-ROTATION ANALYSIS ###################################
# Define two nodes at (0,0)
node 1001 0.0;
node 1002 0.0;

# Fix node 1001
fix 1001 1;
fix 1002 0;

# Define Material Behaviour
set matTag 1;
set s1p 40.e4;
set e1p 1.e-4;
set s2p 50.e4;
set e2p 2.e-4;
set s3p 49.99e4;
set e3p 1.e-3;
uniaxialMaterial Hysteretic $matTag $s1p $e1p $s2p $e2p $s3p $e3p -$s1p -$e1p -$s2p -$e2p -$s3p -$e3p 1. 1. 0.0 0.0 0.0;


# Define element
element zeroLength 2001 1001 1002 -mat $matTag -dir 1;

# Create recorder
recorder Node -file data/stress-strain.out -time -node 1002 -dof 1 disp;

# Define reference moment
pattern Plain 3002 "Linear" {
load 1002 1.0e4;
}

# set maximum Rotation:
set maxR 0.002;
set numIncr 200; # Number of analysis increments to maximum curvature (default=100)
# Compute curvature increment
set dR [expr $maxR/$numIncr];

constraints Plain;
numberer Plain;
system BandGeneral;

test NormDispIncr 1.e-9 50 2;
algorithm KrylovNewton;
analysis Static;


integrator DisplacementControl 1002 1 $dR;
analyze 60;
# for Cyclic Loading
integrator DisplacementControl 1002 1 -$dR;
analyze 120;

integrator DisplacementControl 1002 1 $dR;
analyze 180;
integrator DisplacementControl 1002 1 -$dR;
analyze 230;
integrator DisplacementControl 1002 1 $dR;
analyze 100;

[Horr]

I emphasize "No Peak-Oriented model, Just a trilinear model like bilinear Steel01".

Thank you

[vesna]

You can get a trilinear model by combining two Steel01 materials in parallel. To do that use uniaxialMaterial Parallel: http://opensees.berkeley.edu/wiki/index ... l_Material

[Horr]

That's exactly what I need.
Thank you, very much.