Center of Contact Pressure, getting incorrect results, any ideas why? 1

[MrSamuel]

I am loading a model of a human foot onto a flat plate. There is frictional contact between the bottom surface of the foot's soft tissue and the plate. The foot is about 300mm long (oriented along x axis) and I would think center of pressure would be somewhere in the middle, maybe around 100mm. The back of the heel is located at approx x=0.
I'm using the history output to get 'center of the total force due to contact pressure', specifically, XN1. I get a value of 400. That's 100 forward of the front of the foot. I don't even understand how that's possible...
Any ideas why this is happening and how might I fix it?

Thanks,
Sam

 

[IceBreakerSours]

ABAQUS v6.11 documentation says:

.. center of pressure (XN, XS, and XT) on the surface (defined as the point closest to the centroid of the surface that lies on the line of action of the resultant force for which the resultant moment is minimal. ..

During the application of load the resultant force direction must vary quite a bit. Perhaps, you could apply a dummy no-load step and, in this step, request XN output history variable.

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

 

[MrSamuel]

I created a new step to run after the load step. Loads and boundary conditions were propagated to the new step. Is this what you meant by dummy step? XN1 still was 400, consistently 400 over the entire new step. Any thoughts?

Thanks,
Sam

 

[IceBreakerSours]

Interesting. And I may be wrong.

But, in any case, I meant adding a final step wherein no load is applied. No previous loads need to be propagated.

I am not sure if this "trick" will work at all but I guess if you simply have the previous equilibrium state maintained throughout the dummy step, the force vectors at the contact surface should not change their direction at all and then, I believe, the resultant force vector (for which resultant moment is minimal) should lie somewhere under the foot.

You could also try to request CF/CM (or, perhaps, SOF/SOM) output variables and see what/where the resultant is/applied. You will need to read the documentation for details.

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

 

[MrSamuel]

When I add a new step with no load, the ground and foot separate, so its no longer in the state with which I would like to know center of pressure. Am I missing something?

Oddly enough, the center of pressure abaqus gives is more what I would expect in the middle of this dummy step (~200mm), though the foot is only in contact with the plate at the front, kinda confusing...

Do you think XN1 is really what I want? What do the X and N stand for?

I will look into the other things you mentioned.

Thanks,
Sam

 

[IceBreakerSours]

Very strange; I would not have expected the two bodies to separate. I am not sure about the reasons behind your observations. Sorry.

However, the location of the CoP may not show up where you expect it to be - precisely because of the quotation from the documentation (see my first response).

As far as I can tell, three output variables exist for coordinates of pressure (or, intuitively, normal stress), frictional (i.e., shear stress) and total stress. These are given the identifiers: XN, XS, and XT, respectively. X, as you might guess, is an identifier for coordinate. See section 34.4.1.

Also, why don't you try viewing CPRESS/CSHEAR/CNORMF/CSHEARF (and CPRESSERI/CSHEARERI)? In the Viewer, you could observe how the location of the max. CPRESS or CNORMF moves as the load is applied. In any case, CPRESS/CNORMF may be more relevant to the study.

I made a mistake last time: Section force/moment output requests (SOF/SOM) are not appropriate in this situation.

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

 

[MrSamuel]

refining the mesh a whole lot seems to be helping XN1 take on a more realistic value. Very odd seeing as the pressure distribution doesn't seem to change with refining the mesh, just the center of pressure calculation is no longer bogus...

Sorry for the delayed post and thanks again for all the help,
Sam