Hi all,I have a pressurized piping layout something like this:
-----------
| |
B | |
------------ ---------
| |
| |
| |
__ fixed |
A .
.
.
.
|
|
|
|
C ____ fixed
The model is made of shell181 elements
Kinematic hardening palsticity model
Pipe has internal pressure P.
Analysis type : Nonlinear plastic transient dynamic
analysis under earthquake acceleration
loading and internal pressure.
As can be seen in the figure above ,the pipeing loop has several elbows.
My model has around 66k dof which cant be brought down further.
The model runs but slowly on my 3.02 GHZ ,2GB RAM Pentium IV m/c.
ELbows are the critically stressed elements and I need shell elements on them to see plastic response of elbows.
For reasons of economy of solution,I have tried the model made up of shell elements for a small portion consisting of one elbow and straight pipe portions on both sides of it(A to B in Fig)and modeled the remaining portion of the layout with pipe elements .
The connection between shell model and beam model is imposed by REBAR3 coupling equations.
In the modal analysis the shell+beam mixed model gave almost same frequencies as the full shell model did.
Now in the transient dynamic analysis there is a problem with the shell+beam model.
Due to internal pressure the elbow tries to deflect upward,
giving rise to moment at the fixed end "A" .This moment doent occur in the model consisting of all shell elements,due to fixed condition at "C",the other end of the loop.
The attached pipe element model for remaining portion of the layout is not able to constrain the elbow from deflecting upward due to internal pressure(even though internal pressure is applied in the pipe element model as well)
The loads are internal pressure and earthquake acceleration.
Any suggestions on improving the model for transient dynamic analysis?
-----------
| |
B | |
------------ ---------
| |
| |
| |
__ fixed |
A .
.
.
.
|
|
|
|
C ____ fixed
The model is made of shell181 elements
Kinematic hardening palsticity model
Pipe has internal pressure P.
Analysis type : Nonlinear plastic transient dynamic
analysis under earthquake acceleration
loading and internal pressure.
As can be seen in the figure above ,the pipeing loop has several elbows.
My model has around 66k dof which cant be brought down further.
The model runs but slowly on my 3.02 GHZ ,2GB RAM Pentium IV m/c.
ELbows are the critically stressed elements and I need shell elements on them to see plastic response of elbows.
For reasons of economy of solution,I have tried the model made up of shell elements for a small portion consisting of one elbow and straight pipe portions on both sides of it(A to B in Fig)and modeled the remaining portion of the layout with pipe elements .
The connection between shell model and beam model is imposed by REBAR3 coupling equations.
In the modal analysis the shell+beam mixed model gave almost same frequencies as the full shell model did.
Now in the transient dynamic analysis there is a problem with the shell+beam model.
Due to internal pressure the elbow tries to deflect upward,
giving rise to moment at the fixed end "A" .This moment doent occur in the model consisting of all shell elements,due to fixed condition at "C",the other end of the loop.
The attached pipe element model for remaining portion of the layout is not able to constrain the elbow from deflecting upward due to internal pressure(even though internal pressure is applied in the pipe element model as well)
The loads are internal pressure and earthquake acceleration.
Any suggestions on improving the model for transient dynamic analysis?
