Poroelasticity

[Curt Voss]

I'm been trying out FEBio as a potential new vehicle for our poroelastic cartilage models (currently in Abaqus), especially with an eye toward contact models. I have a few questions I was hoping someone could clarify for me:

--In PostView, is the "pressure" output the fluid pressure, or the hydrostatic pressure component of the solid matrix stress tensor? (-1/3*trace(S))

--What is the default treatment for the fluid phase at free poroelastic surfaces? Is it necessary to apply a 0-pressure nodal constraint to enforce free-draining surfaces?

--Related to the above, how is the fluid flow handled at the contact interface? It appears to be a sealed-at-contact algorithm?

--Finally, in Abaqus it is necessary to apply an initial void ratio, or water content, but I can't find any mention of that here. Is that not necessary for the FEBio formulation?

Thank you in advance for any response. My first impression is that this program is an outstanding advance for poroelastic modeling, and contact specifically.

[maas]

Hi Curt,

The fluid pressure is stored in the temperature field. I know what you're thinking: how could I have not known this ?! Just kidding, the reason for this not so obvious choice is related to the limitations of our current output format (LSDYNA database). We are working on a new output format, which will be more flexible and will allow us to make more accurate descriptions. Also, FYI, the fluid flux is stored in the acceleration field.

Yes, a free draining surface is specified with a 0-pressure constraint.

The contact interface is currently ignored for the fluid phase. (This is a work in progress and we hope to have that implemented in the near future).

Our poroelasticity formulation differs in some regards to ABAQUS's. I would recommend you consult the FEBio Theory Manual (if you have not already done so), where our formulation is described in some detail.

Let me know if you have any other questions.

Cheers,

Steve.

[Curt Voss]

Hi Steve, thanks for the response!

I actually had the thought that I might stash the pore pressure in the temperature field for lack of anywhere else to put it, I had just looked at it wrong and thought that must not be it. The fluid flux field I wouldn't have found though, thanks for pointing that out.

I looked more at your theory manual and see now where the porosity drops out.

Re: Contact -

Your program gives remarkably realistic (intuitive?) results, at least for simple cases of contact between materials with cartilage-like properties. Would it be at all possible to make a dynamically updating set, such that you could prescribe 0 fluid pressure to a set of nodes and at each new increment the nodes that were in contact in the previous increment are removed from the 0 fluid pressure set? Perhaps the full fluid contact algorithm is an easier target than that

I'm hoping to run some 1-to-1 comparisons of poroelastic runs in Abaqus and FEBio. I'll post the results up here when I've done those, if you'd like.

Thanks again,
Curt

[maas]

Hi Curt,

I would love to see comparisons between Abaqus and FEBio. If they compare well, would you mind posting them on the "FEBio Examples and Tutorials" forum instead?

As far as the fluid contact goes, I don't think I will actually implement anything until we have all the theory figured out. It's a complex thing, as you know, but it is also high on our priority list, so hopefully it won't take too long before it finds its way to the code.

Cheers,

Steve.

[michalacadova]

The contact interface is currently ignored for the fluid phase. (This is a work in progress and we hope to have that implemented in the near future).

Hi,
I was wondering how a free-draining surface of a poroelastic material is nowadays handled (using the sliding contact)? When a rigid body moves over this free-draining surface, what happens in the part of this free-draining surface that is in contact? Since in the manual for the biphasic contact is mentioned that “…, fluid can flow from one side of the contact interface to the other.” I would expect that in sliding contact this does not happened.

Michala

[ateshian]

Hi Michala,

We are in the process of reviewing old posts and responding to them, sorry for missing your question.

When a rigid body moves over a biphasic surface, by default there will be no fluid flow into or out of the biphasic material. In other words, rigid bodies are treated as impermeable by default.

If the rigid body is porous and free-draining (i.e., if the fluid pressure in the porous material is ambient), this can be reproduced by applying a zero-pressure boundary conditions on the biphasic contact surface.

So, in summary, rigid bodies may be treated either as impermeable or free-draining.

Best,

Gerard