*SURFACE BEHAVIOR,PRESSURE-OVERCLOSURE=TIED

disla

Hi,

I have followed an interesting video tutorial about how to stablish connection between shells and solids.
I have seen the difference in Stresses continuity between BONDED CONTACT ELASTIC , BONDED CONTACT TIE and CONTACT PRESSURE-OVERCLOSURE=TIED (custom card required).

I have seen MECWAY translates BONDED CONTACT ELASTIC as CONTACT PRESSURE-OVERCLOSURE=TIED, but with a Stiffness equal in both directions and a friction coefficient of 1.
This generates some spurios stresses in the contact area when the materials have different Poisson's Ratio as tangential stiffness “fights” against the material need to accommodate the elemental volumes.

¿Would It be possible to add the PRESSURE-OVERCLOSURE option (Hard, Linear, TIED,..) in future releases with special interest in the =TIED option. It performs very well when one removes the tangential stiffness introduced by MECWAY.

Regards

MECWAY BONDED CONTACT ELASTIC

*CONTACT PAIR,INTERACTION=SI_9,TYPE=SURFACE TO SURFACE,ADJUST=6.454745980652
Shellface3,Solid_Face3
*SURFACE INTERACTION,NAME=SI_9
*SURFACE BEHAVIOR,PRESSURE-OVERCLOSURE=TIED
1E+14
*FRICTION
1,1E+14

CUSTOM DESIREBLE

*CONTACT PAIR,INTERACTION=SI_8,TYPE=SURFACE TO SURFACE
Solid_Face,ShellFace
*SURFACE INTERACTION,NAME=SI_8
*SURFACE BEHAVIOR,PRESSURE-OVERCLOSURE=TIED
1E+14
*FRICTION
0.5

Victor

You want to allow the surfaces to slide against each other but not separate? I think that should be a different type of contact since it has a different physical meaning.

disla

HI Victor,

I would like to be able to set up the *SURFACE BEHAVIOR,PRESSURE-OVERCLOSURE=TIED + the value of the tangential stifness (even remove it). I have not found other option than OVERCLOSURE=LINEAR + automatic tangential stifness but it is too high from my point of view. TIE doesn't work properly.

I'm doing some additional tests and it works really nice. Even with very small contact areas.

-ASTM Pipe 4" Sch10s (114mmx3.01mm) modeled as 2m Solid + 2m Shell and 2m Beam connected in series with *SURFACE BEHAVIOR,PRESSURE-OVERCLOSURE=TIED .
Nice continuity on displacement and stresses.

Victor

If you remove tangential stiffness, then it's not really bonded anymore, is it? It should be free to slide. That might be worth having but I think it's a different thing from bonded contact.

Or are you saying the existing way is producing wrong results? Could you show an example?

disla

The first picture of the post is an example.

Tie and Elastic with high tangential stiffness might warranty the bodies keep together but it can make some stresses to arise at the contact area. Displacements continuities are usually ok.

*SURFACE BEHAVIOR,PRESSURE-OVERCLOSURE=TIED + a user adjustable Tangential Stiffness can help to remove those discontinuities even with dissimilar materials and thickness.

Victor

I think those stress discontinuities are real if they're caused by differential expansion of the two materials, so they shouldn't be hidden by softening the tangential stiffness.

disla

Your comparison between merging the nodes and the elastic contact has convinced me.
Once the stresses arising by an unproper selection of master and slave are removed , the stresses due to different material properties are totally real.

Thanks Victor.