Gay-Berne for discotics

Dear users and developers,
I am running simulations of a Gay-Berne discotic system using lammps. It is stated in the manual
that “It is typical with the Gay-Berne potential to define as the minimum of the 3 shape diameters of the particles involved in an I,I interaction, though this is not required.” However, with reference to Figure 1 of the paper by Bates and Luckhurst J. Chem. Phys. 104, 6696 (1996) doi: 10.1063/1.471387 it seems that the implemented GB model for a disc is what in the paper is called GBDI, Figure 1 (top) in the paper, since by using lammps I obtain pair potential curves identical to model GBDI (in Figure 1top of the paper) and not like Figure 1bottom. In this model, the sigma parameter used to scale and shift the potential is not the minimum axis, but the largest one (for a disc). In contrast, in the model of the paper labeled GBDII and shown in Figure 1(bottom) of the paper, the distance is correctly scaled by the shortest contact distance. However, this seems not to be the case in lammps (contrary to what reported in the manual). Could anyone help? Is there some way to define the input parameters so to have the discotic GB model labeled GBDII in the Bates&Luckhurst paper, where the distance is scaled by the shortest contact distance of a pair of molecules? The definition of the GB potential I use in the input is the following: units lj atom_style ellipsoid dimension 3 read_data config_1.0 set type 1 mass 1.0 set type 1 shape 1.0 1.0 0.345 pair_style gayberne 1.0 2.0 1.0 3.0 pair_coeff 1 1 1.0 1.0 1.0 1.0 5.0 0 0 0 Thank you! –

CCing Mike Brown on this, who may know the answer, as he implemented GB in LAMMPS.


Dear all,
after receiving your email I checked again the manual and I realized
that I made a mistake. Originally, I thought that the sentence in the
manual meant that LAMMPS automatically takes as sigma the shortest
distance of the ellipsoid. I now realize that the sentence, in fact,
simply means the in the literature about Gay Berne, sigma is usually
taken as the shortest axis.
It was my misinterpretation, infact I left in the input file the wrong
value of sigma = 1.0 rather than setting sigma = 0.345 in the line
pair_style gayberne
I am very sorry about this, I hope it did not caused you too much trouble.
Thank you anyway for your help and all the work in mantaining the
software package!
Best regards,

Steve Plimpton <[email protected]> ha scritto: