gran/hertz/history normal dissipation force

Submitted by SergeiD on Sat, 03/24/2012 - 11:46

Hello.
I have a question about the normal dissipation force model for gran/hertz/history pair_style.
From LIGGGHTS documentation and source I see what

F_dissipation = -gamma_n * relative_vel_n

where

gamma_n ~ (S_n*m_eff)^(1/2),
S_n ~ Y_eff * (R_eff*delta_n)^(1/2).

So,

F_dissipaton ~ (Y_eff)^(1/2) * R_eff^(1/4) * delta_n^(1/4).

But as I know, and as follow from ref.[Brilliantov] (see eq.(17)) from the manual, it should be:

F_dis ~ Y_eff * R_eff^(1/2) * delta_n^(1/2).

Whence is this discrepancy?

Forums:

ckloss's picture

ckloss | Wed, 03/28/2012 - 17:25

Hi Sergei,

we are using contact models (Hertz/Mindlin / Linear Spring) which are also used by other commerical DEM codes.
Basically, the damping parameter is derived in a way that the desired coefficient of restitution is pictured.

However, of course there are tons of different models in the literature, you will hardly find two which are the same... We are also well aware that this standard model may produce some unrealistic physics at the particle scale.

A couple of people were interested in this paper
http://www.sciencedirect.com/science/article/pii/S0032591011000325
and started to work on some of these models. if you have suggestions, feel free to bring them forward!

Cheers, Christoph

Somnath | Thu, 03/29/2012 - 00:51

Hi Christoph,

Thank you for the reference to the paper but I don't see the beta parameter in their Hertz-Mindlin model.

beta = ln(e)/sqrt(ln^2(e)+ pi^2)

Could you please guide us to a reference to this model that I can cite for my article?

Thanks
Som