Contact model - hertz vs hertz/stiffness -

Submitted by ykusaka on Mon, 08/06/2018 - 07:37

I'm new to LIGGGHTS and enjoying DEM simulation. I have the following questions.

1. Can I use a cohesion model such as sjkr2 with hertz/stiffness? namely,

fix boxwall all wall/gran model hertz/stiffness tangential history cohesion sjkr2 primitive type 1 xplane 1

is possible?

2. If I use stiffness model with kn, kt, gamman, gammat specified, the values of youngsModulus, poissonsRatio and damping coefficient defined by

fix d1 damp viscous ...

will be ignored? In other words, is it harmful to leave youngsModulus, poissonsRatio and "fix damp viscous .." in my code when using hertz/stiffness?

3. To my understanding, contact models should be defined in both "fix wall/gran" and "pair_style". For example,

fix boxwall all wall/gran model hertz/stiffness tangential history cohesion sjkr2 primitive type 1 zplane 1
pair_style gran model hertz/stiffness tangential history cohesion sjkr2
pair_coeff * *

If contact models defined in "fix wall/gran" and "pair_style" are different, is it harmful?

Thank you for your help.

Forums:

Daniel Queteschiner | Tue, 08/07/2018 - 14:42

1. Yes.
2. fix viscous applies a drag force independent of the contact model, so it will apply a force to the particles no matter what contact model you specify; additional global properties are ignored if not used by the actual contact model.
3. Not only should they be defined in both commands but they must be defined. It rarely makes sense to use different contact models for particle-particle and particle-wall collisions but it should be possible.

RAJ KASHYAP's picture

RAJ KASHYAP | Thu, 08/23/2018 - 12:13

I cannot see any significant change in flow rate by compressing the particles by including both hertz and cohesion model. Which model is suitable for such type of process ?

Thanks in advance

Raj Kashyap