Bridge Deck Modeling issues

[ospinag]

Hello. The code provided below intends to simulate a loaded deck over a bridge. The deck is modeled using shell elements, and the girders are modeled using beam elements. The deck is connected to the girders through rigid links. The load over the deck is 40 kip, and the vertical reactions at the supports should be 10kip at every node. There is something missing here? I can't get the model work, thanks!

## Input code ##
wipe
wipeAnalysis;

set numdof 6;
model basic -ndm 3 -ndf $numdof

set numNode 24;
node 1 -144 -144 0;
node 2 -144 144 0;
node 3 144 -144 0;
node 4 144 144 0;
node 5 -144 -144 9;
node 6 144 -144 9;
node 7 144 144 9;
node 8 -144 144 9;
node 9 -48 -144 9;
node 10 -48 -48 9;
node 11 -144 -48 9;
node 12 -48 48 9;
node 13 -144 48 9;
node 14 -48 144 9;
node 15 48 -144 9;
node 16 48 -48 9;
node 17 48 48 9;
node 18 48 144 9;
node 19 144 -48 9;
node 20 144 48 9;
node 21 -144 -48 0;
node 22 -144 48 0;
node 23 144 -48 0;
node 24 144 48 0;

fix 1 1 1 1 0 0 0;
fix 2 1 1 1 0 0 0;
fix 3 1 1 1 0 0 0;
fix 4 1 1 1 0 0 0;

rigidLink beam 2 8;
rigidLink beam 13 22;
rigidLink beam 11 21;
rigidLink beam 1 5;
rigidLink beam 4 7;
rigidLink beam 20 24;
rigidLink beam 19 23;
rigidLink beam 3 6;

geomTransf Linear 1 -1 0 0
set A_beam_1 3.54;
set Iyy_beam_1 53.8;
set Izz_beam_1 2.18;
set J_beam_1 0.05;
set Say_beam_1 1.39;
set Saz_beam_1 1.88;
set E_beam_1 29000;
set G_beam_1 11153.8;
set Ro_beam_1 7.3446e-07;
element elasticBeamColumn 1 1 21 $A_beam_1 $E_beam_1 $G_beam_1 $J_beam_1 $Iyy_beam_1 $Izz_beam_1 1
element elasticBeamColumn 2 21 22 $A_beam_1 $E_beam_1 $G_beam_1 $J_beam_1 $Iyy_beam_1 $Izz_beam_1 1
element elasticBeamColumn 3 22 2 $A_beam_1 $E_beam_1 $G_beam_1 $J_beam_1 $Iyy_beam_1 $Izz_beam_1 1
element elasticBeamColumn 4 3 23 $A_beam_1 $E_beam_1 $G_beam_1 $J_beam_1 $Iyy_beam_1 $Izz_beam_1 1
element elasticBeamColumn 5 23 24 $A_beam_1 $E_beam_1 $G_beam_1 $J_beam_1 $Iyy_beam_1 $Izz_beam_1 1
element elasticBeamColumn 6 24 4 $A_beam_1 $E_beam_1 $G_beam_1 $J_beam_1 $Iyy_beam_1 $Izz_beam_1 1

set t_shell_1 8;
set E_shell_1 3605;
set G_shell_1 1502.08;
set Ro_shell_1 2.2483e-07;
set Nu_shell_1 0.2;
section ElasticMembranePlateSection 1 $E_shell_1 $Nu_shell_1 $t_shell_1 $Ro_shell_1;
element ShellMITC4 7 5 9 10 11 1
element ShellMITC4 8 11 10 12 13 1
element ShellMITC4 9 13 12 14 8 1
element ShellMITC4 10 9 15 16 10 1
element ShellMITC4 11 10 16 17 12 1
element ShellMITC4 12 12 17 18 14 1
element ShellMITC4 13 15 6 19 16 1
element ShellMITC4 14 16 19 20 17 1
element ShellMITC4 15 17 20 7 18 1

recorder Node -file "supportreactions.txt" -time -node 1 2 3 4 -dof 1 2 3 4 5 6 reaction
timeSeries Constant 1 -factor 1.0
pattern Plain 1 1 {
load 10 0 0 -10.0 0 0 0
load 12 0 0 -10.0 0 0 0
load 16 0 0 -10.0 0 0 0
load 17 0 0 -10.0 0 0 0
}

integrator LoadControl 1.0
test EnergyIncr 1.0e-8 10 0
algorithm Linear
numberer Plain;
constraints Transformation
system BandGeneral;
analysis Static
analyze 1

## Output reactions ##
## 0.4554 5.9337 0.0013 0.3446 -0.0128 -21.9257
## 0.4554 -5.9337 0.0013 -0.3446 -0.0128 21.9257
## -0.4554 5.9337 0.0013 0.3446 0.0128 21.9257
## -0.4554 -5.9337 0.0013 -0.3446 0.0128 -21.9257

[ospinag]

When I change the type of constrain to

equalDOF 2 8 1 2 3 4 5 6
equalDOF 22 13 1 2 3 4 5 6
equalDOF 21 11 1 2 3 4 5 6
equalDOF 1 5 1 2 3 4 5 6
equalDOF 4 7 1 2 3 4 5 6
equalDOF 24 20 1 2 3 4 5 6
equalDOF 23 19 1 2 3 4 5 6
equalDOF 3 6 1 2 3 4 5 6

The output of the nodes 1-8 is

0 0 0.0011 -1.2991 0.0083 0
0 0 0.0011 1.2991 0.0083 0
0 0 0.0011 -1.2991 -0.0083 0
0 0 0.0011 1.2991 -0.0083 0
0 0 9.9989 1.2991 -0.0083 0
0 0 9.9989 1.2991 0.0083 0
0 0 9.9989 -1.2991 0.0083 0
0 0 9.9989 -1.2991 -0.0083 0

By summing the corresponding rows, the resultants are

1+5 0 0 10.0000 0 0 0
2+8 0 0 10.0000 0 0 0
3+6 0 0 10.0000 0 0 0
4+7 0 0 10.0000 0 0 0

which is the result I'm expecting from the beginning. Is that the correct way to get this results from OpenSees? I' would expect to get this directly in the output file, without having to make additional operations.

[vesna]

No, that is not a good way to get the results. To get a reaction at the node you should use reaction recorder at that node.

[ospinag]

Thanks for the reply Vesna. If you mean use "Node recorder -reaction", that doesn't work, because I get the results shown above:

0 0 0.0011 -1.2991 0.0083 0
0 0 0.0011 1.2991 0.0083 0
0 0 0.0011 -1.2991 -0.0083 0
0 0 0.0011 1.2991 -0.0083 0

for the supported nodes. So, how can I get "real" values of nodal forces when using "equalDOF", or "rigidLink" constraints?

Thanks again.

[vesna]

Yes, that is the recorder that you should use. If you are not getting the values you expect there is a problem with your model.

[srishtibanerji]

Hi all
Can I get some examples for bridge modeling?
It will be very very helpful
Thanks
srishtibanerji