Yade

Questions about CohesiveFrictionalContactLaw

Asked by Gastbye Liu on 2018-03-27

Hi everyone,

I am using the CohesiveFrictionalContactLaw and trying to understand how it works.

I noticed:
contactPhysics->normalAdhesion = normalAdhPreCalculated*pow(std::min(geom->radius2, geom->radius1),2);
contactPhysics->shearAdhesion = shearAdhPreCalculated*pow(std::min(geom->radius2, geom->radius1),2);

1) Does that mean normalAdhesion = tensileStrength * pow(min(R1,R2),2) shearAdhesion= shearStrength * pow(min(R1,R2),2)?
But I saw in papers the normalAdhesion should be equal to tensileStrength * pow(min(R1,R2),2) * Pi.
I was wondering that whether I neglect Pi in some place...

2)If I set fragile = true, so the value unp is not used. Right?

Best Wishes,
Liu

Question information

Language:: English Edit question

Status:: Solved

For:: Yade Edit question

Assignee:: No assignee Edit question

Solved by:: Bruno Chareyre

Solved:: 2018-03-27

Last query:: 2018-03-27

Last reply:: 2018-03-27

Link existing bug

Revision history for this message

Gastbye Liu (gastbye) said on 2018-03-27:

And in cohesion model how does the cohesion break?
Through reading the code, I think when cohesion break because of two situlations.
One:
the tensile force between particles exceeds the normal adhesion. Then there would be no force between particles.
Second:
The shear force exceed the shear adhesion and there would be no force between particles.

Am I right?
Thanks!

Revision history for this message

Bruno Chareyre (bruno-chareyre) said on 2018-03-27:

Hi Liu,

1) Yes

> But I saw in papers the normalAdhesion should be equal to ...

I would say that if you try to justify why it "should" be something you'll end up realizing that it can be whatever you like, and "should" not be anything specific. As soon as you know which equations you are using you should be fine.

2) not sure now. I have the impression that it should be used anyway, but you better try by yourself.

3) yes

Bruno

Revision history for this message

Gastbye Liu (gastbye) said on 2018-03-27:

Thanks for your reply!

By the way, in CohesiveFrictionalContactLaw, does it consider the the moment caused by tangential force?

Revision history for this message

Bruno Chareyre (bruno-chareyre) said on 2018-03-27:

I would say yes but... which moment?
B

Revision history for this message

Gastbye Liu (gastbye) said on 2018-03-27:

In https://github.com/yade/trunk/blob/master/pkg/dem/CohesiveFrictionalContactLaw.cpp
I can see:
Vector3r moment = phys->moment_twist + phys->moment_bending;

But when calculating moment, the tangential force would affect moment. M = Fs * R.
So... I thought in CohesiveFrictionalContactLaw, it doesn't consider the moment Fs * R ?

Revision history for this message

Bruno Chareyre (bruno-chareyre) said on 2018-03-27:

The line you mention defines the contact moments (bending,twisting) independently of the shear force.
Shear force gives a torque on the body, not a contact moment.
B