Building a simulation of a multimaterial object

[Paolo-Susini]

Hi all,

I'm currently working on building a simulation environment where an anatomical finger collides with soft materials or objects.

To make the simulation as realistic as possible, I’d like the finger to be composed of multiple distinct layers, such as:

• The outer skin layer (with appropriate material properties)
• The dermis
• The subcutaneous tissue
• The bones

To be more specific, I already have FEM models for each individual part — skin, dermis, subcutaneous tissue, and bones — and each of them contains internal cavities where the adjacent inner tissue is meant to be embedded.

I'm currently trying to use HexahedronCompositeFEMForceFieldAndMass, but I'm having a hard time implementing it. Collisions between an external object (e.g., a point mass) and the composite material don’t result in any deformation of the composite, as also mentioned here.

I also came across this example, but I'm not sure whether it's suitable for my use case.

[hugtalbot]

Hi @Paolo-Susini

Thank you for this interesting question.
As usual with FEM, several approaches could be considered.

If you go for a 3D modeling, your finger could define different zones with various elasticity properties. An example of this approach is the one you referred: Heterogeneous-TetrahedronFEMForceField.scn. It is well suited if each zone is distinct from another.

You could go for a mapping-based or a rigidification approach, especially for the bone model if it is immersed in the volume of a 3D skin model.

Some companies around SOFA also implemented 3D layered (i.e. surface) mechanical models, but these models are not in the open-source project.

I hope this sheds some light on your simulation problem.
Best

[Paolo-Susini]

Hi @hugtalbot,

To give some context, I’m attaching the following images:

These represent two FEM models I’m using to simulate a real human finger. One has the internal layers physically separated (e.g., bone, muscle, skin), while the other is a unified volume.

I’d like your opinion on the best modeling approach going forward:

Would it be better to:

Use a single filled FEM model, and manually assign different material properties (e.g., Young’s modulus, stiffness) to different elements inside the mesh,
or

Use the layered/segmented model, where each anatomical layer is a distinct mesh region?

Is there a known best practice for this kind of multi-material biological simulation? I’d appreciate any advice or workarounds you can recommend based on your experience.

Thanks a lot in advance!

Best regards,
Paolo

[hugtalbot]

Hi @Paolo-Susini

Sorry for my latency.

Different approaches could be tested. A single FEM model indeed, but I fear that since you have some bones, the Young's modulus difference might be important between the different finger parts. Thus, leading to ill-conditioning of the system matrix.
I would rather go for a mix with a single FEM for the muscle, skin (for now) and define some rigification for the bones. Later, a separate skin layer (surface mechanical model) could be added using a mapping.

Hope this helps

[github-actions[bot]]

⚠️ ⚠️ ⚠️
@Paolo-Susini
Feedback has been given to you by the project reviewers, however we have not received a response from you. Without further news in the coming weeks, this discussion will be automatically closed in order to keep this forum clean and fresh 🌱 Thank you for your understanding

[github-actions[bot]]

⚠️ ⚠️ ⚠️
@Paolo-Susini
In accordance with our forum management policy, the last reply is more than 4 months old and is therefore closed. Our objective is to keep the forum up to date and offer the best support experience.

Please feel free to reopen it if the topic is still active by providing us with an update. Please feel free to open a new thread at any time - we'll be happy to help where and when we can.