Restart

[Chris74]

Hi all,
I have a simple question about loading to a restart from the solution.
First, I have mechanical results from a previous compute (strain,stress,dispacement), carried out in step 1.
Second, I'd like to restart from the last step of these results with new boundaries : no specific loading, only new blocked boundaries:
My questions are :
- are the stresses in the structural part considered as loading ?
- should I apply a loading as a force field ?
Thank you for your advice concerning this post.

[kevinarden]

Chapter 3
https://www.nic.funet.fi/pub/sci/physic ... Manual.pdf

You have to tell Elmer to save the restart file in the first analysis.

I have done this with transient, in 1 solution, have not tried it in restart.
In transient I apply the load for x amount of time, then let it go.

viewtopic.php?p=20818&hilit=abaqus#p20818

[Chris74]

Thank you Kevin for your reply.
In the past, I've done a study following your suggestion (loading for a x acount of time and let it go), and I works well. I could see the damping of the structural part.
But now the problem is different, I will do it whitout restart like this :
0 <time < t :
- loading on boundary 1 (Dirichlet conditions, position of the nodes)
- fixed at 0 on boundary 2

time >= t :
- fixed on boundary 1 (strain position at time = t)
- free on boundary 2

does this look ok?

[Chris74]

I was thinking of changing my Dirichlet conditions with the transient, but I don't see how to do it.

I want to block nodes for a period of time and release the blocks afterwards. In other words, I want to apply 1 boundary condition only for the first steps.

Boundary Condition 1
Target Boundaries(2) = 18 85
Name = "BoundaryCondition 1"
Displacement 1 = Variable "time"
Real MATC "if (tx<=0.8) 0.0;"
Displacement 2 = Variable "time"
Real MATC "if (tx<=0.8) 0.0;"
End

=> it doesn't work because there is no value for the displacement variables after t >0.8.

How can I do ?
Should I go into restart mode with new boundaries conditions ?

[kevinarden]

I would have to test if doing by restart allows you to change boundary conditions.

Otherwise there are conditional boundary conditions

Displacement 1 = Real 0.0
Displacement 1 Condition = Real $condx

When $condx > 0, the condition is enforced, when $condx < 0, not

So perhaps

Displacement 1 = Real 0.0
Displacement 1 Condition = Variable Time
Real MATC "if (tx<=0.8) 1.0; else -1.0;

"

[kevinarden]

I redid the transient damping test with displacement BC instead of load force, and the results look reasonable. it vibrates after being let go.

vibration.png

(129.57 KiB) Not downloaded yet

damped.sif

(2.61 KiB) Downloaded 10 times

[Chris74]

Thank you very much Kevin, I will test it tomorrow !