Soils - Contact - ABAQUS CAE

[Gregorio Marchiori]

Good morning,

I am having a lot of troubles in converging a simulation with a lot of contacts, large deformations, permeability and soils steps. The input file is attached. If you have some suggestions, you are wellcome.

In particular I'm not confident with soil steps in this context.

Thank you very much in advance.

 

[FEA way]

Can you describe how this analysis should work (what should happen in each step) ?

That’s the knee joint model, right ? It consists of two disks representing bone ends with meniscus between them. Let’s focus on the mechanics. In the first step (Contact) there’s a pulling force applied to the top disk but also displacement pushing it down. Meniscus is fixed via connectors and the bottom disk moves upwards. Then in the next step (Pre-Load) meniscus is still fixed and bottom disk moves upwards while top one is pushed down with concentrated force. Last step (Stance) consists of pretty much the same BCs as previous one (apart from the top disk that has different rotational DOFs constrained) and uses different force magnitude.

Is this how the model should work ? What physical process is it simulating ?

 

[Gregorio Marchiori]

Yes, it is a "Knee" phantom. The two disks are the cartilages of femur (above) and tibia (below), respectively. First step (contact) is displacement controlled to reach the starting pose (presence of the force is an error), tibial cartilage moves upwards, femural cartilage downward. meniscus in the middel. Second step is load controlled, to give a pre-load. Third step should be the full load in a walking stance. All the steps are soils, but only the third is transient to activate the fluid flow. If you look at the time period of the third step, 0.125 s, it is our time frame in the gait analysis.

 

[FEA way]

There are many negative eigenvalue warnings in this analysis. This suggest some kind of instability. I think that the mesh should be denser. Try adding automatic stabilization in step settings and using general contact instead of contact pairs. Consider adjusting material properties or using hyperelastic model.

 

[Gregorio Marchiori]

I'm tring to reproduce what is explained in a scientific article, in terms of material properties, mesh density and contat definition. Using general contact is helpful, but not enough. I tried automatic stabilization, but solution converged only with too high dumping

 

[FEA way]

Can you share the title of this article ? Maybe something in your model is slightly different than described in the paper causing convergence issues.

 

[Gregorio Marchiori]

I have also the more complex knee model, similar to the articles, but now I'm tring with the simplified model I attached here. The articles are:

"Comparison between kinetic and kinetic-kinematic driven knee joint finite element models"

"Identification of locations susceptible to osteoarthritis in patients with anterior cruciate ligament reconstruction: Combining knee joint computational modelling with follow-up T1ρ and T2 imaging wiet" with supplementary material.

 

[FEA way]

Convergence issues are mentioned in both articles. In the first one authors inform that they had to adjust parameters such as contact, mesh and solver settings. You should try the same. Your model diverges at some point close to the end of the analysis when the deformation becomes too large. And it seems that there is some penetration.

 

[Gregorio Marchiori]

There is a lot of penetration. I have built the semplified model exactly to investigate the key points on contact and solver settings that those authors don't reveal, but it is a big challenge to me

 

[Gregorio Marchiori]

From the literature references above, could you be able to suggest the kind of Steps with some relative specifications?