Holzapfel-Gasser-Ogden material for arteries

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • ateshian
    Developer
    • Dec 2007
    • 1824

    #16
    Hi,

    The answer to your question depends very much on the geometry you are analyzing and the material properties you are using, in relation to the 15 kPa pressure you are applying. If you have tried to adapt the model given at the beginning of this thread, by only changing the axial prestrain to 10% and an internal pressure of 15 kPa, the analysis should run to completion, see the attached file ( SilviaR.feb ).

    Best,

    Gerard

    Comment

    • maas
      Lead Code Developer
      • Nov 2007
      • 3400

      #17
      Hi Silvia R ,

      It is possible that the model is somewhat underconstrained or that the applied load is too large. For a static analysis this could cause problems. In a dynamic model, the material's inertia is taken into account, which often will prevent the mesh from deforming drastically.

      If the contact is penetrating, you may need to increase the penalty parameter (which essentially scales the applied contact force). In addition, you can turn on the auto-penalty flag, which tries to estimate a good initial penalty parameter. In that case, the penalty parameter will scale this initial auto-penalty value.

      If this does not help, please attach or send us the model file, so we can take a closer look and perhaps offer additional suggestions.

      Best,

      Steve
      Department of Bioengineering, University of Utah
      Scientific Computing and Imaging institute, University of Utah

      Comment

      • emmieschoutens
        Junior Member
        • Apr 2021
        • 1

        #18
        Hello,

        In the new version of FeBio (1.4), there is a material possibility of the HGO unconstrained model implemented. I did simulations on the geometry of an abdominal aneurysm. I do think I use normal values for all input parameters and I have tried several options on the solver, including the allowed residual. However, the simulation keeps running into divering or negative Jacobian problems. The .feb file was too large to upload, so in the attachments I have included the input parameters for the HGO unconstrained model and an overview of the used mesh.

        Is there any advice in how to prevent these errors from happening?

        Thanks in advance!

        Attached Files

        Comment

        • ateshian
          Developer
          • Dec 2007
          • 1824

          #19
          Hi,

          I checked the values you are using for the "HGO unconstrained" material model and there is nothing wrong with them, as far as I can tell. If you are having difficulties with your model, it could be due to the mesh, or boundary conditions, or the magnitudes of loads you are applying?

          Let me know if you are still encountering problems with this model.

          Best,

          Gerard

          Comment

          • mkgale123
            Junior Member
            • Jun 2021
            • 10

            #20
            Originally posted by ateshian View Post
            Hi,

            Please don't use that model, it doesn't work. It was included in FEBioStudio by mistake. We will take out for the release version of FEBioStudio.

            Best,

            Gerard
            Hello,
            Is this still the case in the current version of FEBio Studio? The same model with the same parameters is still available under 'uncoupled elastic' in the materials list.

            Comment

            • ateshian
              Developer
              • Dec 2007
              • 1824

              #21
              Hi,
              The current HGO models are working, please try them out.
              Best,
              Gerard

              Comment

              • mkgale123
                Junior Member
                • Jun 2021
                • 10

                #22
                Hi Gerard,

                I have some questions about the appropriate material axes to use. I'm attempting to generate an HGO-like material model but with 4 fiber classes (one parallel to the circumferential direction, one parallel to the axial direction, and two at equal but opposite angles alpha from the axial direction) and use this to model a uni-layer hollow cylinder. When I use the <1,4,5> axis set (as demonstrated in the tutorial at the top of this thread), I find that the resultant model experiences some twisting along along the axial direction in a way that I don't believe is correct. When I use the default <1,2,4> axis set, this twisting does not occur, but I'm not certain if my desired fiber angles are then correct because I'm not familiar enough with material axes. Would you help me determine the appropriate axis set? I've attached a link to an upload of my working model if that would help.

                Thank you,
                MK

                Comment

                • ateshian
                  Developer
                  • Dec 2007
                  • 1824

                  #23
                  Hi MK,

                  You can add the following plot variables (manually in the <plotfile> section of the .feb file) to your model:
                  Code:
                  <var type="material axes"/>
                  <var type="fiber vector['solid[1]']=fiber 1"/>
                  <var type="fiber vector['solid[2]']=fiber 2"/>
                  <var type="fiber vector['solid[3]']=fiber 3"/>
                  <var type="fiber vector['solid[4]']=fiber 4"/>
                  After you solve the problem, you can display the material axes (determined by <mat_axis type="local">1,4,5</mat_axis>) using the Vector Plot option, then selecting which axis to display in the Data Field. This will show you how the local element axes (1, 2, 3) are oriented in your model.

                  Solids in a solid mixture are internally numbered starting from 0. Since the four fibers in your solid mixture represent the second to fifth <solid> entries, you can display their fiber directions using solid[1] through solid[4] (the first solid in the mixture is solid[0]). The labels "fiber 1" etc. are up to you (you can call each fiber anything you want). Each of these fibers can be displayed using a Vector Plot and selecting the corresponding Data Field. You can display multiple fiber vectors together by creating multiple Vector Plots.

                  Please try it out and let me know if that addresses your concern.

                  Best,

                  Gerard

                  Comment

                  • mkgale123
                    Junior Member
                    • Jun 2021
                    • 10

                    #24
                    Gerard,

                    That allows me to see exactly what I need. Thank you!

                    Best,
                    MK

                    Comment

                    • razaghisahar@yahoo.com
                      Junior Member
                      • Feb 2023
                      • 1

                      #25

                      Hello Steve and Gerard,




                      Again, I am sorry for bothering you with my questions.




                      I have checked the forum, but I have faced multiple questions, and I would appreciate it if you could assist me with those.




                      1- I checked the stl (surface import) and stp (volume import) formats from the SolidWorks model to FEBio. However, when I select the stl option, the curve option cannot directly choose the boundary. The other problem is that the shape is an unusual format imported into FEBio (some parts of the surface cannot be meshed using the mesh section of FEBio). In contrast, in the case of the STP format, the shape has a 3D dimension. I want to consider a 2D surface of a skin model that is uniaxially being displaced along x axis. In this case, selecting a curve is not working, and only selecting the surface works, which differs from what I am looking for.




                      2- How can I select the boundary by selecting the curve after mesh refinement? (it seems the edge should be manually defined instead of using a select curve)




                      3-Is it possible to define using FEBio quad and tri mesh elements simultaneously without using any interface such as GMSH? What kind of software are you using for meshing?




                      4- What is the equivalent of the Gasser Ogden Holzapfel model in Ansys, Abaqus or COMSOL?




                      5- I am defining the model through only script; how would I know that the unit is SI? For instance, I am using these values for the corresponding paper (https://doi.org/10.1098/rsif.2005.0073). Is that correct?

                      Is the model defined by default by Pa, MPa, or GPa?




                      <k>100</k>

                      <pressure_model>Abaqus (GOH)</pressure_model>

                      <c>0.1007</c>

                      <k1>24.53</k1>

                      <k2>0.1327</k2>

                      <kappa>0.1456</kappa>

                      <gamma>90</gamma>




                      6- Is the GOH model only the one mentioned above? What if I want to increase the number of fibres?

                      Is it correct to say that if the fibres' orientation is aligned with displacement, the gamma is 0 degrees, and if the fibres' orientation is perpendicular to displacement, the gamma will be 90 degrees? Is gamma based on degree or rad? You mentioned on your website that it is based on rad. Am I right?




                      7- According to my model, if I want to stretch the skin along the x direction within a stretch ratio of 50 percentage, the value of the displacement boundary should be 12 (for x=24 initial dimension).- according to my model, if I want to stretch the skin along x direction within the stretch ratio of 50 percentage, the value of displacement boundary for should be 12 (for x=24 initial dimension)?




                      8 —If the bulk modulus is higher in the GOH model, the model will be stiffer; how should I define this value?




                      9 —How can I use partition meshing in your software? Is there a tutorial on that? I want to consider quad mesh for the whole surface and be closer to the skin model. He also mentioned that only quad mesh gives rise to a soft response, and considering both meshes gives me a more realistic answer for the skin model.




                      10—Could you please assist me in understanding how the control setting parameters in the static and dynamic explicit models work and what each parameter defines?







                      Thank you for all your assistance in advance and best regards,

                      Sahar





                      Comment

                      Working...
                      X
                      😀
                      😂
                      🥰
                      😘
                      🤢
                      😎
                      😞
                      😡
                      👍
                      👎