Yade

Use of JCFpmPhys.dilation ?

Asked by Jérôme Duriez

Hello,

I was wondering if JCFpmPhys.dilation was used in the computations with JCFPM model ? (It seemed to me, not) Or if it is just for post-processing ?
Especially, is there really, in the current source, a dilatancy term in the computation of the normal force ?

Thank you !

Jerome

Question information

Language:
English Edit question
Status:
Solved
For:
Yade Edit question
Assignee:
No assignee Edit question
Solved by:
Luc Scholtès
Solved:
Last query:
Last reply:
Revision history for this message
Best Luc Scholtès (luc) said : 		#1

Hi Jerome,

Good to hear from you.

You are right, I have not used dilation in any modelling but you have to note that it is just a variable that does not enter in any computations, unlike dilationAngle. I put it there because I thought it could be interesting to quantify dilation in joints and, actually, after double checking, I am pretty sure it is not computed the right way (see line 112 of the cpp file). Thanks for pointing out. I will have a look.

Coming back to dilationAngle, it has been used in computations (see figure 8 in this paper: http://www.sciencedirect.com/science/article/pii/S1365160912000391) and, it is working when the smoothJoint contact logic is turned on. Implementation is also on line 112 of the cpp file and has been verified. There could be a discussion on the way it is implemented and I am keen for remarks.

To sum up:

- Yes, a parameter is currently present and functional in the source code to take into account dilatancy in joints, it is dilationAngle.
- dilation is a variable that should record the amount of dilatancy in the joint but needs to be corrected.

Hope it helps

Luc

Revision history for this message
Jérôme Duriez (jduriez) said : 		#2

Ok ! I did not catch the change in "initD", I was looking rather to a change directly in Fn...

One extra remark about this topic : I thought that it would maybe sound more "rheological" to take into account dilation only at plastic sliding, and, then, to compute the normal dilation according to the plastic part of the relative shear displacement. Do you agree ?

Once I'll do calculations, I'll maybe try to code that. Not sure however that it changes anything (in the results)...

Revision history for this message
Jérôme Duriez (jduriez) said : 		#3

Thanks Luc Scholtès, that solved my question.

Revision history for this message
Luc Scholtès (luc) said : 		#4

"One extra remark about this topic : I thought that it would maybe sound more "rheological" to take into account dilation only at plastic sliding, and, then, to compute the normal dilation according to the plastic part of the relative shear displacement. Do you agree ?"

Actually, if you have a look at the code, dilation (the change in initD) is taken into account only when maximum shear strength is reached (plastic yield) and it takes into acount only the incremental shear displacement (jointSliding).

Now, definitely, the computation of the normal displacement in the joint (phys->dilation), needs to be corrected. What do you think about that way:

phys->dilation += jointSliding*phys->dilationAngle

So we use the incremental shear displacement (after plastic yield)?

Luc

Revision history for this message
Jérôme Duriez (jduriez) said : 		#5

Yes, for me it looks ok !

(excepted that this "jointSliding" occuring during one time step has itself probably an elastic and plastic part, but taking this into account is "cutting the hair into 4")