O.cell.vellgrad simple shear test

Asked by jsonscript on 2020-07-17

Hello everyone:

I want to ask the mechanism of O. Cell.Velgrad, What is the loading mechanism of O. Cell.Velgrad, and if it is used to simulate the simple shear test(like O.cell.velGrad=Matrix3(0,0,0, 0,0,0, 0.02,0,0)), is it reasonable?

Also there is another way to simulate the simple shear test.which the sample is sandwiched between two plates..the top plate was secured to keep the initial height of specimen and simple shear was applied by moving the bottom plate.

What are the differences between the above two methods of simulating a simple shear test..

Thanks!

Question information

Language:
English Edit question
Status:
Solved
For:
Yade Edit question
Assignee:
No assignee Edit question
Solved by:
Jan Stránský
Solved:
2020-07-20
Last query:
2020-07-20
Last reply:
2020-07-17
Jan Stránský (honzik) said : #1

Hello,

> I want to ask the mechanism of O. Cell.Velgrad, What is the loading mechanism of O. Cell.Velgrad,

O.cell.velGrad (not Cell and not Velgrad) is velocity gradient of the periodic cell (see e.g. [1])
If this is not your answer, please describe more what is "mechanism" and "loading mechanism".

> if it is used to simulate the simple shear test(like O.cell.velGrad=Matrix3(0,0,0, 0,0,0, 0.02,0,0)), is it reasonable?

yes

> Also there is another way ...

there are many other ways

> What are the differences between the above two methods of simulating a simple shear test..

Is this a question? If yes, please put question mark "?" at the end of questions.

cheers
Jan

[1] https://www.continuummechanics.org/velocitygradient.html

jsonscript (jsonscript) said : #2

Sorry , I am apologizing for my poor English..
I am trying to calibrate torsional shear experimental results and want to use periodic boundaries on a RVE to replicate the shearing process. Is it reasonable to use periodic simple shear case?

Also there is another way to simulate the simple shear test.which the sample is sandwiched between two plates..the top plate was secured to keep the initial height of specimen and simple shear was applied by moving the bottom plate.

What are the differences between the above two methods of simulating a simple shear test?

Jan Stránský (honzik) said : #3

> I am apologizing for my poor English..

your English is good, the information provided are sometimes problematic :-)

> I am trying to calibrate torsional shear experimental results and want to use periodic boundaries on a RVE to replicate the shearing process. Is it reasonable to use periodic simple shear case?

sounds reasonable, but depends on the definition of "torsional shear experiment".
What is the shape of the specimen?
What material is it (mainly cohesive or not)?
What is the loading.
...

> What are the differences between the above two methods of simulating a simple shear test?

You have already answered this, or? :-) The main difference is the loading:
Periodic:
- is periodic / does not use walls
Walls:
- is not periodic / does use walls

cheers
Jan

Hello Jan and Jsonscript,

I am also interested in this question.

If I understand correctly in periodic simple shear, by using O.cell.velgrad, all the particles of the assembly are assigned a velocity based on their position with reference to the cell. Which mimics the continuum shear case. Where deformation is linear, as usually observed in a continuum body.

But, in a hollow torsional cylinder torsion shear tests on sands, the deformation may not be linear, right? As we apply Torque on specimen using a shaft on the top end of the specimen while bottom end is fixed. Is it still reasonable to consider periodic simple shear scenario in replicating the element stresses of a torsional shear experiment?

I am assuming Jsonscript, might be talking about "hollow cylinder torsion shear tests on sands".

Thank you
Vivek

Jan Stránský (honzik) said : #5

Hello,

> by using O.cell.velgrad, all the particles of the assembly are assigned a velocity based on their position with reference to the cell.

roughly.
(by default) velocity of particles is composed of this "macroscopic" or "homogeneous" term ("a velocity based on their position with reference to the cell") and their own interactions with real and "periodic" members.

> in a ... cylinder torsion shear tests on sands, the deformation may not be linear, right?

yes, it may be non-linear. That's why I asked for more information about the experiment.

> Is it still reasonable to consider periodic simple shear scenario in replicating the element stresses of a torsional shear experiment?

as usually, it depends on many factors :-)
- what is the dimension of periodic cell compared to the specimen?
- Is the axial dimension fixed, or is the specimen free to dilate in axial direction?*
- what are the "trnasverse boundary conditions"?*
- periodicity also forces strain localization (shear bands) to be periodic. Is it a problem or not?
- ...

* in "true simple shear", all "normal" dimensions of the cell are fixed, only the shear is applied. Depending on the experiment, it may not correspond to reality and some adjustment may be needed (e.g. zero strain in one direction but zero stress in a perpendicular direction).

cheers
Jan

Hello Jan,

Thank you for your advise.

>(e.g. zero strain in one direction but zero stress in a perpendicular direction)
I didn't quite understand this adjustment you talked about. You mean we make some adjustments in the boundary stresses and strains to replicate the experiments, without necessarily adding any wall type boundaries?

Thank you
Vivek

Best Jan Stránský (honzik) said : #7

> ... without necessarily adding any wall type boundaries?

it is one of the main reasons to use "periodic boundary conditions", not to add wall type boundaries

> You mean we make some adjustments in the boundary stresses and strains to replicate the experiments

yes. I mean just that the situation in the experiment may be different from O.cell.velGrad=Matrix3(0,0,0, 0,0,0, 0.02,0,0))

cheers
Jan

jsonscript (jsonscript) said : #8

Thanks Jan Stránský, that solved my question.