Convergence Problem with displacement Control

[arashs]

Hi,
I'm working on a bridge model and I need to verify my material models for different components. I'm using a 2DOF truss element to verify my Uniaxial material models, some of which include an initial gap. I've tried displacement control, load control and Arc length methods to test those materials with the initial gap but they all fail in the first increment. It is obviously because of the initial gap ( where tangent line to load/displacement curve becomes horizontal) and I have no how to solve that. Anyone with the same problem???? here is my code:
###############################################################
wipe;
model BasicBuilder -ndm 1 -ndf 1
# Define nodes
node 1 0.0
node 2 1.0


# Fix node 1
fix 1 1


######### Pounding (impact) element material model--------------------------
set kt1 1116e6
set kt2 384e6
set initialgap 0.6
set yielddisp 2.54e-3
# uniaxialMaterial ImpactMaterial $matTag $K1 $K2 $deltay $gap
uniaxialMaterial ImpactMaterial 1 $kt1 $kt2 -$yielddisp -$initialgap


####---------------------------------------------------------------------------------------

# Define truss element with unit area
# tag ndI ndJ A matTag
element truss 1 1 2 1 1

########-----------------------------------

# recorders
recorder Node -file disp.out -time -node 2 -dof 1 disp
recorder Node -file force.out -time -node 1 -dof 1 reaction



pattern Plain 1 Linear {
load 2 30.0e5
}


constraints Transformation
numberer RCM
system BandGeneral
test NormDispIncr 1.0e-8 10
algorithm Newton

###### Cyclic loading
foreach Dincr {-0.660 0.66} {
integrator DisplacementControl 2 1 $Dincr
#integrator ArcLength 0.0001 10
analysis Static
analyze 4000
}

[brag006]

You could probably add another element with a very low elastic modulus. It will ensure the analysis proceeds and wont affect your results

[arashs]

brag006 wrote:
> You could probably add another element with a very low elastic modulus. It
> will ensure the analysis proceeds and wont affect your results

Thanks for your hint, but how is it gonna help???? should I run the analysis using displacement control?? if so how should I differentiate displacement from two elements???

[fmk]

you could put the material in parallel with another material AND FROM YOUR KNOWLEDGE ABOUT THE BEHAVIOUR OF HOW 2 SPRINGS WORK WHEN IN PARALLEL you could deduce if the material is working or not!

failing that you could look at the example on the following gap materials page to see another way of testing uniaixal materials:
http://opensees.berkeley.edu/wiki/index ... p_Material