Yade

Trying to understand simple shear in 3D periodic systems

Asked by Steven Tobin

Hello all,

I am currently looking at studying shear bands in granular materials, and have started working with Yade to that end.

Building from the periodic simple shear example (which uses velGrad to deform the cell), I currently have simulations that seem to lead to sensible stress/strain behaviour and shear band formation. I am only shearing in one direction.

How can I conceptually reconcile a velocity gradient across the sample with triply periodic boundaries? I have no issue with the boundaries perpendicular to the shear direction (front to back, side to side), but the parallel (top to bottom) can't wrap my head around. Or, frankly, am I completely wrong in my approach here?

Question information

Language:
English Edit question
Status:
Solved
For:
Yade Edit question
Assignee:
No assignee Edit question
Solved by:
Bruno Chareyre
Solved:
Last query:
Last reply:
Revision history for this message
Best Bruno Chareyre (bruno-chareyre) said : 		#1

On my side I'm trying to understand why top-bottom direction is more
confusing for you than the other directions. ;)
Check the yellow boxes in [1], they show a collection of contiguous
simple-sheared periods. What is the problem?

[1]
https://yade-dem.org/doc/formulation.html#approximate-collision-detection

Revision history for this message
Steven Tobin (steventtobin) said : 		#2

Thanks for replying so quickly Dr Chareyre. The figure you linked to clears up my confusion!

In the 2D example on that page my issue could have been phrased as: a particle may wrap in x (side-side) direction with no issue, but if the particles at the bottom are moving much slower than the particles at the top due to the presence of the shear band, how can the top-bottom boundary condition be satisfied? The figure highlights the relative motions of the "periodic cells" wrt one another - the missing piece for both me and a colleague I posed the question to.

Revision history for this message
Steven Tobin (steventtobin) said : 		#3

Thanks Bruno Chareyre, that solved my question.