How to solve the re...
 
Notifications
Clear all

How to solve the re-oscillation problem of gravity in full restart analysis?

4 Posts
2 Users
1 Likes
653 Views
Posts: 3
Topic starter
(@evanyi)
Barista
Joined: 3 years ago

Dear everyone:

I am trying to preload an RC specimen with gravity. In order to achieve that, I employed the mass weight method to apply gravity. The model and corresponding artificial damping curve are shown in Fig.1 and Fig.2.
After the structures achieve static equilibrium in the first analysis (Fig 3), I proceed into the full restart analysis stage. The dump01 file of the first analysis is copied into a new fold for restart analysis, as well as the original keyword file with only modification on CONTROL_TERMINATION and STRESS_INITIALIZATION (To check if the gravity is applied properly).
However, here the problem comes. After the restart analysis finished, the contact force showed a kind of high-frequency-oscillation, as shown in Fig 4. I also checked axial force in the cross-section of the middle beam, similar problems also exist.
Relevant introducing documents about preload or restart only provide brief roadmaps to these techniques, while solutions for this kind of detail problems are still lacking as much as I can recall. So, I hope anyone who used to encounter similar problems could provide solutions for this problem. This is precious for me and any future researchers concerning preload and restart problems.
Thanks for your attention.
3 Replies




Negative Volume
Posts: 639
Admin
(@negativevolume)
CEO
Joined: 5 years ago

Hi @evanyi

Now this is very interesting. I'm going to do some digging on this and get back to you because I've always thought that restart simulations can provide a lot of value but never quite work the way I intend them to. 

But please anyone else comment if you can help @evanyi

Reply
1 Reply
(@evanyi)
Joined: 3 years ago

Barista
Posts: 3

Thank you, @negativevolume

Hope you can find out some enlightening things. I turned to another traditional but less efficient way to solve my problems at present, which is combining the gravity equilibrium and the impact process into a transient analysis. It did avoid the re-oscillation problem and is employed by many researchers in their papers as I can recall, but it is pretty time-consuming.

A lot of thanks again.

Reply




Posts: 3
Topic starter
(@evanyi)
Barista
Joined: 3 years ago

Recent Update:

FYI, this problem had some interesting updates recently. Some people who encountered the exact problem contacted me and said that he found out it has something to do with the interaction between the beam and solid elements. If the model is simulated with homogeneous elements, for example, plain concrete only, then the restart analysis could be fulfilled without this tricky problem. Some enlightening clues.

Reply







Share: