Integrated Finite Element Model in Adams

Question

Dear community,

I have a multibody dynamics model of human body in Adams. Currently the intervertebral discs (IVD) are modeled as bushings with stiffness and damping. I am trying to model IVDs with flexible bodies option in Adams. I was able to import a Wavefront (.obj) file of the IVD and use Rigid to Flex option to convert it to a flexible body. I have a couple of questions regarding the modeling options:

1. I created two fixed joints between markers on the IVD and the inferior and superior bodies. But I want to tie the whole surface instead of just one point. What is the best option for this?

2. The material properties of IVD components are defined with hyperelastic Mooney–Rivlin model. How do I define hyperelasic material properties in Adams?

3. I am modeling the IVD as one whole geometry, but IVD is composed of three main components, nucleus, annulus, and collagen fibers. The nucleus is usually defined as incompressible fluid, the annulus is defined as 5-10 concentric layers, and the fibers are modeled as springs. What is the best tool to model each component? Is Adams a good option to do this or I need to use another software. I appreciate any suggestions.

4. How do the different components of a flexible body interact with each other in Adams? In other words, is it feasible to model all components as one flexible body and expect to see their interaction?

Alexis Del Rio · Accepted Answer

Hi Hani,

As already mentioned this quickly becomes a non-trivial piece of work. If I were you I would isolate a pair of vertebrae with the disc in between and test/learn. About your questions:

Test different options; RBE2, RBE3 nodes implementation then solid to solid, flex to flex contact... and try to understand which one makes more sense based on experimental results. The most complex, time consuming simulation may not be the one that provides the best results 
 if you are not confident with subroutines you may want to try using state variables... to create your own 'Mooney–Rivlin' mathematical model 
 again state variables... may be a good starting approach here. While you may want to cosimulate with Marc, the question here is 'How non-linear is this elastic component?' can you live with a simplification? 
 This is one type of model that is not that easy to figure out, hence the need to test different things. My expectation is that you may not need a full flexible vertebrae (??) maybe you can create a sandwich with a couple of flex discs top/bottom, a rigid body in between and two fixed joints. Then work out the contact between this flex disc and the IVD. I would import the geometry in MSC Apex and prepare the geometry there first, it is way easier to do there.