PPGap-pushover test problem

[mpaksoy]

Dear Vesna,

I was trying to model the shear keys on a bridge bent. A super small part of code is below. My assumption is to bond, bearing material in transverse direction and also the shear key effect as parallel materials. By the way for sure I use displacement control PO. In ElasticPPgap material we can assign the behaviour only in 1 direction + or negative. As a result of this shear key effect on rubber bearing will only be in one direction. I noticed this when I run cyclic pushover!! Everything is fine in 1 directional pushover. However, when I later run NLTHA, I have to get in both positive and negative response.
How can I define my spring representing elastomer bearing and shear key effect in both directions. Should I define zerolength elements for both directions individually. Would you please give me a hand about the problem? Thank you very much.

Murathan

set shearkeygap 0.03
#BEARING PROPERTIES
set Fybtrans 245.0; #YIELD STRENGTH FOR A TOTAL NUMBER OF 8 NEOPRENE
set Fyblong 245.0
set btransstiff 35000.0; #(KN/M)
set blongstiff 35000.0; #(KN/M)
set bvertstiff 1000000000.0; #(KN/M)
set botherstiff 1000000000.0;


#DEFINITION OF BEARING Materials
uniaxialMaterial Steel01 $bmattrans $Fybtrans $btransstiff 0.00001
uniaxialMaterial Steel01 $bmatlong $Fyblong $blongstiff 0.00001
uniaxialMaterial Elastic $bmatvert $bvertstiff
uniaxialMaterial Elastic $bmatother $botherstiff

#gap closing effect and Bearings
uniaxialMaterial ElasticPPGap $skeymat [expr 150000*$Fybtrans] [expr -15*$Fybtrans] [expr - $shearkeygap]
uniaxialMaterial Parallel $parallelmat $bmattrans $skeymat
element zeroLength 200 200 201 -mat $bmatvert $parallelmat $bmatlong -dir 1 2 3 -orient 0 1 0 0 0 -1;#bearing connecting the bent to previous span
element zeroLength 202 202 203 -mat $bmatvert $parallelmat $bmatlong -dir 1 2 3 -orient 0 1 0 0 0 -1;#bearing connecting the bent to next span

[vesna]

Yes, you need to define two gap materials; one for each direction.