No announcement yet.

viscoelastic UMAT

  • Filter
  • Time
  • Show
Clear All
new posts

  • viscoelastic UMAT

    Dear members,

    I am a master student. For my master thesis I need to develop a UMAT code for epoxy resin to simulate the residual stress during curing. The problem is that i am an abaqus beginner. If some of you have a viscoelastic UMAT code with Kelvin elements or Maxwell, can he sends it to me please.
    PS: I have tested the Abaqus documentation UMAT example but there are 3 parameters which are not defined.

    Thank you in advance

  • #2
    old thread


    • #3

      Thank you Frank. I checked you thesis but even if you use 6 parameters, your model is different since it is a Maxwell well one. I have to deal with the Kelvin model and for this one, the others 3 parameters are unexplained.

      Do you have an idea?

      Thank you


      • #4
        Did you note there are 2 pdfs in the old thread ?


        • #5
          Originally posted by Frank View Post
          Did you note there are 2 pdfs in the old thread ?
          I have checked the latter pdf file. But not only it is german but the Kkv Gke etc are not defined first. Do you know what it represent ?


          • #6
            The terms denote:
            GKe = shear modulus of the elastic component in the Kelvin body
            KKv = bulk modulus of the viscous component in the Kelvin body
            GE = shear modulus of the single spring under the Kelvin body
            and so on
            Check my thesis on the viscous Poisson's ratio. Commonly it is taken as 0.5, making Kkv infinite.

            The sketch on page 4 shows a "standard solid". If you want to simplify it to a Kelvin body then make the elastic properties of the single spring labeled "E" infinite.
            This way the left-hand side in the last equation on page 5 leaves only Sigma_ij, the right-hand side remains unchanged. This should be a Kelvin body in 3d.

            The equation used in my thesis is for a Zener body, see page 3. In your model you arrive at a product KKv*(time derivative of the trace of the strain tensor). Probably the latter must be set to zero, and infinity multiplied by zero is zero.


            • #7
              Thank you a lot Frank. It seems more clear in my head now. I think with this explaination and you thesis I will do it .