Elastic-Perfectly Plastic Gap Material

[Layla]

Dear Vesna,

I'm trying to simulate the boundary conditions at the deck ends of a bridge and I want to consider the embankment flexibility.
I used an elastomericBearing element for the bearing and a zerolength element with the Elastic-Perfectly Plastic Gap Material for the gap between the deck and the abutment (I assigned small values for the stiffness because I ignore the pounding between the deck and the abutment).These two elements are connected in parallel.
Then, for the embankment, I want to use a material that is perfectly plastic in compression and doesn't have tensile strength, so I used again the Elastic-Perfectly Plastic Gap Material (but I assigned a very small value for the gap).
The problem is that I get errors..
Could you please help me?
Here is a part of my code..
Thank you in advance,
Layla

# left deck end
node 1001 0.00 1.20 -2.60;
node 1003 0.00 1.20 -2.60;
node 1004 0.00 1.20 -2.60;
fix 1004 1 1 1 1 1 1;
# -----------------------------------------------------------------------------------------
# 9. BEARINGS
# -----------------------------------------------------------------------------------------
# build bearing material
set kInit 9112;
set fy 76.540;
set alpha 0.5;

uniaxialMaterial Elastic 100 2700889; # vertical stiff
uniaxialMaterial Elastic 101 11812; # rotational stiff
uniaxialMaterial Elastic 102 7146; # rotational stiff
uniaxialMaterial Elastic 103 10; # torsional stiff

# bearings at the abutments
# element elastomericBearing $eleTag $iNode $jNode $kInit $fy $alpha -P $matTag -T $matTag -My $matTag -Mz
element elastomericBearing 102 1003 1001 $kInit $fy $alpha -P 100 -T 103 -My 101 -Mz 102 -orient 0 1 0 1 0 0;

# -----------------------------------------------------------------------------------------
# 10. GAPS - ABUTMENTS
# -----------------------------------------------------------------------------------------
set matGapx 140;
set Kabx [expr 0.0000000001]; # NO POUNDING
set Fpoundx [expr -10000]; #
set gapx -0.10;

set matGapy 150;
set Kaby [expr 0.0000000001]; # NO POUNDING
set Fpoundy [expr -10000]; #
set gapy -0.15;

# uniaxialMaterial ElasticPPGap $matTag $E $Fy $gap <$eta> <damage>
uniaxialMaterial ElasticPPGap $matGapx $Kabx $Fpoundx $gapx; # long direction, gap
uniaxialMaterial ElasticPPGap $matGapy $Kaby $Fpoundy $gapy; # trans direction, gap

# gaps
element zeroLength 106 1003 1001 -mat $matGapx $matGapy -dir 1 3;

# Abutment material
set matAbx 160;
set Eabutx 338000;
set epsyPx 0.017;
set Fabutx -5624;
set pseudogapx -0.00000000010;

set matAby 170;
set Eabuty 100000;
set epsyPy 0.017;
set Fabuty -1666;
set pseudogapy -0.00000000010;

uniaxialMaterial ElasticPPGap $matAbx $Eabutx $Fabutx $pseudogapx; # long direction
uniaxialMaterial ElasticPPGap $matAby $Eabuty $Fabuty $pseudogapy; # trans direction

set matTagElastic 91;
set Elastic 1000000;
uniaxialMaterial Elastic $matTagElastic $Elastic;

# Abutment
element zeroLength 110 1004 1003 -mat $matAbx $matTagElastic $matAby $matTagElastic $matTagElastic $matTagElastic -dir 1 2 3 4 5 6;

[vesna]

Would you post the errors that you get?

[Layla]

Thank you very much for your reply. I get the following message:

WARNING: CTestEnergyIncr::test<> - failed to converge after 1000 iterations
ModifiedNewton::solveCurrentStep<> - the ConvergenceTest object failed in test <>
DirectIntegrationAnalysis::analyze<> - the Algorithm failed at time 4.449
OpenSees> analyze failed, returned: -3 error flag -3

Moreover i'd like to add that if I plot the analyses results (long displacement) from 0 to 4.449 seconds for the bearing element and the deck end node, they are not similar. The deck response seems reasonable but the bearing's not..
Thank you very much for your help!
Layla

[vesna]

What you have here is a convergence problem. To overcome it you need to do analysis in a loop similar to this example: http://opensees.berkeley.edu/wiki/index ... e_Analysis

[Layla]

Thank you Vesna.
If I use the code lines that change the algorithm whenever is necessary than the analysis may be complete. The problem is that I don't take reasonable results..
Could you please tell me if the connections among the zerolength elements and the elastomeribearing are correct?
I mean is it ok to create more than one zerolength elements between 2 coincident nodes?
In my case, I used the same nodes 1001 and 1003 for the gap and the bearing and then 1003 and 1004 for the abutment (all these nodes 1001, 1003, 1004 are coincident).
Many thanks in advance,
Layla

[vesna]

Yes, you can have two zeroLenght elements between two coincident nodes.

To figure out what is wrong with your model I suggest you to build your model in stages. Start with the two nodes and elements between them and check the results. Add the third node and other elements and check the results.

[clairelove]

Hi,Layla
I'm also using PPGap material and truss for the gap between the decks in the bridge. And I want to get the pounding effects. But i feel confused to set the gap value and the original distance value between to nodes. Are they the same value or different? And how to set the value $fy properly? My code:
# nodes of the link element
node 1 0 0 0
node 2 $distance 0 0
uniaxialMaterial ElasticPPGap 1111 2E8 -$fy -$gap
element truss 1 1 2 5.044 1111
just note the $distance -$fy -$gap value. I don't know the material well.
Thanks!