home.social

Kevin Moerman 🔓🦿:julia:

View on fosstodon.org
  1. Working on nested triply periodic lattices. Here a Gyroid levelset image is created and two interwoven/complementary surfaces are obtained by constructing isosurfaces for the same positive (white) or negative level (red). In the animation the levelset level is adjusted so the surfaces "shrink away" from each other.

    More on Gyroids: en.wikipedia.org/wiki/Gyroid)

    Coming soon to

  2. Playing with of an "expansion mesh" like structure. The geometry, mesh and boundary condition creation, as well as the FEA analysis, all occur automatically.

    This simulation is still elastic, next step is to add plasticity.

  3. Working on extruding or "thickening" meshes from surface elements to layers of solid elements. Here the test mesh features quads (left) which are thickened to produce hexahedral elements (right).

    This is especially handy when one wants a high quality structured mesh of a thin structure. For instance meshing of blood vessel like structures.

  4. Went with another classic, the dragon statue.

    This shows triangulated surface remeshing using a (very basic) wrapper for @BrunoLevy01 et al.'s fantastic Geogram library (github.com/BrunoLevy/geogram).

  5. Coming soon to : constrained triangulations. Which I decided needs parameterized curves too for testing purposes 🦇

  6. So satisfying to finally reach 100% code coverage for testing!

    Hoping to publish paper on Comodo by the end of the year. If you are into help to contribute to Comodo.jl and FEBio.jl and you'll be one of the authors!

    github.com/COMODO-research/Com

  7. Interesting paper by @jedbrown et al.

    doi.org/10.48550/arXiv.2401.13

    For computational mechanics/physics, if you code by just punching in the equations from the textbooks directly, the physics should work, but computationally the way you evaluate the quantities may be unstable. This paper lists some recipes to avoid these.

    Mostly small strain problem, but still feels icky to leave in.

    @mofem @likask @koehlerson

  8. Just read about "rabbit functions", a term for math expressions that are essentially "pulled out of a hat"! Love it, and have to use this in a paper some day.

    From "Wachspress 1975, A Rational Finite Element Basis:

    "A "rabbit function" is a function that is pulled out of a hat. When a magician pulls a rabbit out of a hat, the effect astonishes the onlookers. Once the illusion is revealed, the rabbit generation is
    no more surprising than the generation of rabbits by rabbits."

  9. New : "Experimental and Computational Analysis of Energy Absorption Characteristics of Three Biomimetic Lattice Structures Under Compression" arxiv.org/abs/2308.14452

  10. Added textured OBJ model import to GIBBON. The OBJ import supports mixed faces (e.g. tri+quads).
    The demo video shows coupled refinement of both the geometry and texture data for an imported model.

    Also useful for resampling digital image correlation data on 3D surfaces.

  11. Saw a paper on a soft robotic star thing, and had to check. Yep I can confirm, it looks like it can grab things.

    Paper: doi.org/10.1002/aisy.202200435

    Here is a link to my fully parameterized implemention with + : gibboncode.org/html/DEMO_febio

  12. GIBBON is ready for ! Heading to Paris tomorrow and can't wait to present the latest implementations and upgrades for FEBio4.

    If you are attending the workshop, upgrade GIBBON to the latest version:
    gibboncode.org/

    cmbbe-symposium.com/2023/works

  13. In case you want your tongue Model to go eeeewwww. This GIBBON demo features active contraction of mapped "muscle fibers".

    I just updated it for FEBio4.

    gibboncode.org/html/DEMO_febio

  14. Checking out "Taubin" smoothing. Looks great for smoothing voxel derived meshes. Video shows voxel mesh (red) of a femur and the Taubin smoothed surface retrieved (white).

    One of Taubin's papers: doi.org/10.1109/ICCV.1995.4668

    GIBBON implementation: gibboncode.org/html/HELP_smoot

  15. Here the strain that you think you are applying (red curve) and the real strain in the middle layer (green curve) are plotted over time. As you can see a compressive Z-strain builds up during clamping which may offset the strain graph a lot.
    Note this is Green-Lagrange strain. The linear tensile strain intended to be applied is 30%, the linear clamping strain is -30%.

    gibboncode.org/html/DEMO_febio

  16. Indentation with a twist.

    The rigid ball pushes into the rubber-like hyperelastic cube while also spinning. Friction then transfers torsional forces to the cube as well.


    gibboncode.org/html/DEMO_febio

  17. New with collaborators Shaiv Parikh et al.

    "Biomechanical characterisation of thoracic ascending aorta with preserved pre-stresses"

    doi.org/10.1101/2022.12.02.518

    Features some lovely automated meshing and with GIBBON

  18. A ridiculous model for whoever wants to simulate a rubbery propeller hitting a stick.

    I just used this to play with a contact algorithm and was curious to see if it would work.

    : gibboncode.org/html/DEMO_febio

  19. Here is a nice review paper on: "What Kelvin might have written about Elasticity"
    doi.org/10.1111/j.1365-2478.20

    And more on benefits for use with :
    arxiv.org/pdf/1605.09606

  20. A simple demo for automated design, meshing and of a hyperelastic tire in contact with a road surface.

    GIBBON implementation: gibboncode.org/html/DEMO_febio

  21. Lovely new paper by PhD student I co-supervise: Mahtab Vafaeefar (Twitter: @MahtabVfa)

    Have a look if you are into: lattice structures, trabecular bone, and/or

    "A Morphological, Topological and Mechanical Investigation of Gyroid, Spinodoid and Dual-Lattice Algorithms as Structural Models of Trabecular Bone"

    link
    arxiv.org/abs/2211.13036

    Journal link
    doi.org/10.1016/j.jmbbm.2022.1

    Uses GIBBON codes for lattice structure+model creation

  22. Automated meshing of a bifurcation (e.g. for blood vessel ) with hexahedral elements using toolbox.

    1) Take a cylinder
    2) Cut a hole at bifurcation curve
    3) Create direction vectors for surface "departure and arrival"
    4) Use smooth bezier spline based lofting
    5) Thicken quads to form hex. elements.

    Demo rendered here:
    gibboncode.org/html/DEMO_mesh_

    Bezier approach was inspired by @acegikmo 's spline work