soft tissue incompressibility; uanisohyper_inv 1

[IceBreakerSours]

Dear all,

I am trying to add a pair of incompressible soft tissues in 3D to an anatomical structure. One of the ends of each of the soft tissue meshes (C3D8H) is tied to a stiff anatomical surface and the other end is in hard contact with a block that is constrained to be rigid. The reference node of the rigid block is coupled to another node where the rest of the structure is loaded with a pure moment.

The problem that I am facing is that the soft tissue near the contact surface flattens/thins out which results in the pair of soft tissue meshes penetrating each other. The issue (flattening out) is demonstrated here: Link

Any suggestions will be highly appreciated.

Thanks.

~Ice!

 

[rstupplebeen]

Please attach the picture on this site. The site you pointed to is not very good. Below your response window there is an attachments section.

Rob Stupplebeen

https://sites.google.com/site/robertkstupplebeen/

 

[IceBreakerSours]

Thanks Rob. The image is properly attached now. If you'd like to see another view, please let me know.

 

[IceBreakerSours]

Sorry, I am still trying to get the hang of it.

 

[IceBreakerSours]

Attached is another snapshot. Image on the left shows the thickness direction. The pattern of deformation, I guess, suggests its either the anisotropy/fiber orientation or incompressibility that might be causing the issue. If so, perhaps, increasing the number of elements in the thickness direction might be of some help?

I recognize the other problem is slippage (of the top surface of the soft tissue along contact tangential direction) out of contact. This issue, however, is easy to fix.

Thanks.

 

[IceBreakerSours]

Dear group,

I will deeply appreciate any advice on this issue; I have been struggling with this for a month now!

Thanks.

 

[rstupplebeen]

Could you more clearly explain what is the issue or post your CAE file? I am having a hard time interpreting your images.

Rob Stupplebeen

https://sites.google.com/site/robertkstupplebeen/

 

[IceBreakerSours]

Let me give it another try. If this does not work, I will find out if I can upload the CAE.

I have a pair of abdominal muscles (rectus) in a 3D model of human spine. The bottom element surface of the muscles are tied to the pelvis. The trouble arises at the top surface. In the FE model, there is no rib cage, no thoracic region, no neck/head, etc. The goal is to add rectus muscles to the FE model of the spine.

For visualization, see, for example, here: here (scroll down to the rectus abdominis part).

I have tried the following:

a) Kinematic coupling between the top surface and a node at the top of the spine where the spine is loaded.

I tried this with C3D8H and C3D20H elements. C3D20H elements generated zero pivots/overconstraints, whereas C3D8H elements did fine except the flattening out issue as described previously.

b) It was suggested to me that the "flattening" out was due to incompressibility (because the nodes of the elements are constrained to keep the volume of each element in the muscle constant). Therefore, allowing the nodes at the top surface to move in the tangential direction should help. So, I introduced an interaction with a rigid surface (a block in the pictures above) near the top. The rigid surface was coupled to the node loading the spine.

This did not resolve the "flattening" issue either.

c) Currently, I am adding more elements in the thickness direction to see if that will do the trick.

Again, it is the flattening/thinning out of the elements closer to the top surface of the muscle, and consequently, the element penetration that is the problem I am facing.

I hope this explains the problem a little better.

http://www.eng-tips.com/faqs.cfm?fid=376

 

[rstupplebeen]

Unfortunately, I do not see any clear direction that I can provide with the information you have provided. I could ask a million questions like are you using Standard or Explicit..... If you can't post your model can you create a simplified model that does not have any proprietary information? When creating models I typically start with very simple geometry and assumptions and add complexity from there. The added benefit is that the proprietary information is only added at the end. I hope to be more help in the future. Thank you.


Rob Stupplebeen

https://sites.google.com/site/robertkstupplebeen/

 

[IceBreakerSours]

Thank you very much for trying and I apologize for the inconvenience. Once I know my supervisor is fine with this, I will upload the CAE immediately. I do not think it should be an issue but I want to be sure.

If it is any help, I am using Standard and performing a static analysis; pure moment is applied at the top of the spine to bend it backwards. I have also attached a zoomed-out image of the model.

http://www.eng-tips.com/faqs.cfm?fid=376

 

[rstupplebeen]

Ice,
Here is a list of things that I would check:
1. The material properties of the abs since you call out that they are anisotropic. Poisson's ratio can get tricky.
2. The material orientation of the abs
3. If you are not concerned with the loading at the abs's cut then I would just fix the end instead of using a tie constraint
4. Is there a reason you are not using symmetry?
I hope this helps.

Rob Stupplebeen

https://sites.google.com/site/robertkstupplebeen/

 

[IceBreakerSours]

Hi Rob,

Thanks a lot for the suggestions.

1. Material model has been verified/validated.
2. Orientation of fibers is along the length of the abs and is duly incorporated.
3. I can not fix the end (instead of tie) because loading in to the pelvis is important. I could, however, introduce an interaction instead of tie.
4. Yes. In future, I plan on apply "complex" loads and introduce "imperfections" in to the muscles in an asymmetric fashion.

My only option left now is to see if increasing the number of elements along the thickness of the abs will help in allowing the nodes to move more freely instead of forcing elements to thin out.

http://www.eng-tips.com/faqs.cfm?fid=376

 

[IceBreakerSours]

Hi Rob,

Could you please suggest another option for me to be able to upload the CAE for you? Every time I attach the CAE, a Runtime Error message shows up. I doubt if its the size of my CAE file; its less than 5 MB.

~!ce.

http://www.eng-tips.com/faqs.cfm?fid=376

 

[rstupplebeen]

Sorry for the slow reply, life happens. If you brows to my webpage you can email me directly.

Rob Stupplebeen

https://sites.google.com/site/robertkstupplebeen/

 

[IceBreakerSours]

Thanks Rob. But I think I've found the source of my problem. It's the sharp transition in the stress-strain curve (active muscle fiber property). As you know, numerical algorithms have a tough time at sharp transitions. Trusses defined with no compression/tension behavior end up running into convergence issues too. Unfortunately, stabilization is not ideal. So, I am trying to see if adding and then deleting some connector elements is possible. This way I can provide some stiffness to the material near that transition point and then get rid of it.

http://www.eng-tips.com/faqs.cfm?fid=376