Results 1 to 6 of 6

Thread: Simulate lung deflation with Ogden hyperelastic material

  1. #1
    Join Date
    Jul 2020
    Posts
    7

    Default Simulate lung deflation with Ogden hyperelastic material

    Dear all

    I have been simulating lung deflation successfully with an hyperelastic material (ogden) with the commercial software
    Altair Radioss (see first attached image for illustration)
    there 4 components:
    - one shell mesh that represents a rigid cavity in which the lung deforms
    - a tetra 4 nodes mesh for the lung
    - 2 tetra 4 nodes mesh for the airway exterior and interior
    the boundary conditions are the following:
    - set of fixed nodes with no displacements
    - set of nodes with gravity applied (lung, airway)
    the loads are:
    - pressure on the half top part of the lung with patient on back (use a defined subpart of the lung exterior surface)
    see 2nd attached image for illustration
    a sliding contact between the cavity and the lung

    third attached image is the visualization of my rewriting of this finite element model in the febio format (work in progress)
    i tried to upload the febio file but it is too big
    What is your preference to send a link to big files in this forum?

    i have several questions:
    - i put all the component meshes in the same object, the lung and airways have points in common. Will the contact be handled by defaults in this case?
    - when i describe several parts in an object, how do you recommend to increment the element id? shall there be a global element number in the multi part object? or shall i restart the increment at 1 for each part?
    - for my pressure load definition shall i define a surface section in the geometry section or a surface data in mesh data
    - for the sliding contact do you recommend sliding facet-on-facet or node facet ?

    thank you
    regards
    Anne-Cecile Lesage
    Attached Images Attached Images

  2. #2
    Join Date
    Nov 2007
    Location
    Salt Lake City, UTAH
    Posts
    2,329

    Default

    Hi Anne-Cecile,

    Regarding sharing large models, I recommend using some type of file sharing service such as Dropbox or Google drive. If you prefer not to make your model publicly available, you may also email it to us. However, perhaps try the following suggestions regarding your other questions first:

    - i put all the component meshes in the same object, the lung and airways have points in common. Will the contact be handled by defaults in this case?
    Contact is not handled automatically, so you will need to connect the different components that have separate meshes. If the meshes are conforming at their interface (the interface nodes of both meshes coincide), then the best option is just to merge the meshes. Alternatively, you could add a tied contact interface, but that would add a perhaps unnecessary complexity to your model.

    - when i describe several parts in an object, how do you recommend to increment the element id? shall there be a global element number in the multi part object? or shall i restart the increment at 1 for each part?
    If you bring in your models into FEBio Studio, you should not have to worry about element numbering. The element IDs are generated when you write your model to .feb file.

    - for my pressure load definition shall i define a surface section in the geometry section or a surface data in mesh data
    I assume you are talking about the feb input file? In the input file there will a Surface definition that defines the surface that the pressure should be applied to. A surface data section would only be necessary when you want to apply position-dependent surface pressures.

    - for the sliding contact do you recommend sliding facet-on-facet or node facet ?
    Of the two you mention, I would recommend the "sliding facet-on-facet", but I would suggest to try "sliding-elastic" interface first. This is non-symmetric formulation (the "sliding facet-on-facet" is a symmetric formulation), so you will need to set the "Matrix symmetry" flag to "Non-symmetric". You can find this flag under the Step properties.

    I hope this will help, but let us know if you have any further questions.

    Best,

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

  3. #3
    Join Date
    Jul 2020
    Posts
    7

    Default

    Dear Dr Maas

    thank you for your answers

    First i would like to clarify that i am not writing the .feb files with the studio but with a self-written code that rewrites the fem in radioss format in the xml format of the .feb files
    So being carefully with elements numbering in differents parts of an object might well be necessary in the algorithm of my code, isn't it?
    Ihen i do not want to merge the parts because i want to be able to put different materials for each, the lung properties vary with distance to the main airways
    I guess then i will need to write contacts between my parts, isn't it?

    My project with febio is to reproduce my results with Radioss and then to move to poroelastic material which from a biological point of view is a more suitable material to simulate lung deflation
    than the ogden material i use in radioss.
    The problem is that radioss does not handle well poroelastic for soft tissues. This is why i am trying to use febio instead to pursue my research on lung deflation modeling to the next step.
    the research work is in second review step in Physics for Medecine and Biology. I hope it will be published soon. I hope to improve the accuracy with poroelastic modeling.

    I will continue to work on my fem in the feb format with your recommendation and send it to you privately for more expert recommendation.

    thank you
    Regards
    Anne-Cecile

  4. #4
    Join Date
    Nov 2007
    Location
    Salt Lake City, UTAH
    Posts
    2,329

    Default

    Hi Anne-Cecile,

    If you are writing your own code for generating the .feb file, then you have to indeed ensure that each element gets a unique id.

    Merging the objects in FEBio Studio preserves the partitioning, so you should still be able to assign different materials after merging. Even more, you can always re-partition the mesh after merging to make any necessary changes to the partitioning. If that is your only concern, I would still recommend to merge the parts and avoid the contact. Of course, this is assuming that the interface nodes of both parts coincide.

    Let us know if we can offer any additional help for converting your model to FEBio. Feel free to reach out via email as well (steve dot maas at utah dot edu).

    Cheers,

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

  5. #5
    Join Date
    Jul 2020
    Posts
    7

    Default

    Dear Dr Maas,
    Did you receive en email from me on thursday at your utah.edu adress. just checking it did not go in your junk email box.
    regards
    Anne-Cecile

  6. #6
    Join Date
    Nov 2007
    Location
    Salt Lake City, UTAH
    Posts
    2,329

    Default

    Hi Anne-Cecile,

    I did receive the email, but I did not get a change to reply to it. I'll try to reply to it within the next few days.

    Cheers,

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

Posting Permissions

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