Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations MintJulep on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Convergence difficulties composite shells

Status
Not open for further replies.

TrygveNils

Civil/Environmental
Joined
Mar 6, 2015
Messages
3
Location
NO
I am modeling a hexagonal system with shell elements, see added image. These shell elements are assigned a composite section with 2 materials (linearly elastic) and three layers. I am trying to compress this system transversely with a displacement controlled analytical rigid plane, but the analysis "never" converges. The bottom edge is locked with y-symmetry and the side edges are constrained from moving sideways. I want the system to buckle completely and be squeezed so much it only has about 1/4 of its height left. Any tips on how to improve convergence?

Things i've tried:
Tried both Penalty and Augmented lagrange as interraction property between analytical rigid plane and shell elements
Make the walls thinner (worked to some degree, but I need them to be quite thick)
Set poisson's ratio almost to zero
Fiddle around with general solution controls
Change mesh. (With S3 elements abaqus compresses the system much more than S4. With finer mesh, convergence decreases!)

 
 http://files.engineering.com/getfile.aspx?folder=9159601e-b625-43c9-815f-84fd0ef71546&file=Uten_navn.png
Try an explicit simulation they work much better for buckling problems.
 
Thank you for your answer, I'm looking into it now. I need to know that the results I'm getting truly are quasi-static. I was thinking if the kinetic energy of the system is much less than the internal energy of the system, then the results should be realistic. Is it that simple?
 
That depends on the kind of simulation you are doing.

An explicit simulation will give a dynamic response so you will get oscillation (ie if you are using purely linear elastic materials there is no reason why these oscillations should dampen out when using displacement control) & not a quasi-static response.


 
IceBreakerSours, what do you mean by application=quasi-static? There is already some numerical damping present (bulk viscosity) to damp out the unwanted displacement.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top