Results 1 to 4 of 4

Thread: How much may "Element Distortion" and "Negative Eigenvalue" affect the accuracy?

  1. #1

    Question How much may "Element Distortion" and "Negative Eigenvalue" affect the accuracy?

    Hi
    Dear All,

    I use Abaqus/Standard. Element distortion occurs frequently in my analysis and apparently it doesn't matter how hard I try to change the meshes , at least a few number of elements would be distorted. However, If you create an acceptable meshing , the solution will converge and the job will be completed but with a number of warnings indicating excessive distortion of some elements.

    I am really eager to know to what extend may the existence of distorted elements adversely affect the accuracy of the results , specially the displacement field?

    The same for negative eigenvalues. This warning seems to happen without any special clue , I think it indicates the possibility of buckling if the load is exerted in the opposite direction, so I think this is not a negative factor to the accuracy of my analysis, is it?

    I am very thankful for your insightful replies
    Yours Sincerely,
    Bio

  2. #2
    Join Date
    2000-02
    Location
    Boston, USA
    Posts
    3,328
    When elements get distorted they become less accurate, and may crash a simulation. It is quite possible that the predicted displacements, etc, are not very good in regions with high element distortions.

    When this happens it is best to work out a new mesh, or remesh.

    -Jorgen
    Jorgen Bergstrom, Ph.D.
    PolymerFEM Administrator

  3. #3
    Thanks Dr. Bergstrom,

    Is it possible for viscoelastic creep response not to reach the equilibrium in a given time because of element distortions?
    Since I'm working with Standard , I guess my only option is using mesh to mesh solution mapping and you know, this is a painstaking procedure and can be very exhausting specially when you have to run the model over and over again.

    Because of the considerable soft matter deformation in my model , I feel that however I mesh the instance , it comes time increments in which a group of elements are going to be distorted, of course I can give it a good meshing and prevent it from being failed but I can not eliminate all of those distortions. Do you find my justification to be right or you think there is always a better way of meshing (not adaptives) to do this right?

    ALE adaptive meshing is also not a choice in /standard. DO you see any other way? what do you mean by remesh?
    Do you think Hypermesh software (Altair) can mesh the instance better than ABAQUS?

    Thank you very much and my apologies for the multiple questions,
    Yours Sincerely,
    Bio

  4. #4
    Join Date
    2000-02
    Location
    Boston, USA
    Posts
    3,328
    It is difficult to answer your questions in general.
    Can you provide a more specific example with figures of the mesh in order to illlustrate the element distortions?

    -Jorgen
    Jorgen Bergstrom, Ph.D.
    PolymerFEM Administrator

Similar Threads

  1. Problem about "Large volume changes with geometric nonlinearity"
    By willness in forum Constitutive Models
    Replies: 1
    Last Post: 2009-02-20, 05:13
  2. "total mechanical strain" in ANSYS
    By ANSYSuser in forum Computer Software
    Replies: 2
    Last Post: 2008-06-23, 15:34
  3. Accuracy of Elastomer Models
    By Jorgen in forum PolymerFEM News
    Replies: 0
    Last Post: 2007-08-28, 21:10
  4. Please help me with "Prony Series"
    By survesh in forum Thermoplastics
    Replies: 1
    Last Post: 2006-05-29, 19:48
  5. Negative Eigenvalues
    By anantha in forum Finite Element Modeling
    Replies: 1
    Last Post: 2005-02-23, 19:15

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •