Hello Community,
I have a question regarding the implementation of the contact models. Like I understand and has been discussed allready in the forum Liggghts follows the integral implementation of the spring force according Di Renzo and Di Margio. Thats fine. I am more unsure about the implementation of the damping force and the coulomb criterion.
Like i understood from literature there are 2 ways:
1) First check tangential spring forces for coloumb considering the normal spring force and than add the damper afterwards in both cases if necessary:
F_t_spring < mu * F_n_spring --> F_t_sum = F_t_spring + F_t_damp ; F_n_sum = F_n_spring + F_n_damp
F_t_spring > mu * F_n_spring --> F_t_sum = mu * F_n_spring; F_n_sum = F_n_spring + F_n_damp
In this case the idea is to treat the spring forces as contact forces and the damping is just necessary to avoid "rattling" following the idea of Cundall and Strack.
------------------------------------------
2) Check total tangential forces for coloumb considering the total normal force including dampers:
F_t_spring + F_t_damp < mu * (F_n_spring + F_n_damp ) --> F_t_sum = F_t_spring + F_t_damp
F_t_spring + F_t_damp > mu * (F_n_spring + F_n_damp) --> F_t_sum = mu * (F_n_spring + F_n_damp)
In this case the total forces are considered more follwing implementation like Walton and Thorton.
-------------------------
In LIGGGHTS however the tangential spring force (F_t_spring) is checked against the total Normal force mu *(F_n_spring + F_n_damp). For me this looks like a mix up of both assumptions? Or I just dont get it.
Can anyone explain the reason for the implementation or have any additional sources for the implementation?
Thanks, Thomas