Dear lammps users,
I would like to model the Ti/HfO2 interface, and I have some question about how to develop the parameters, in particular for Ti-Hf:
-
how are the cutoff R and D chosen? Are they pair dependent?For example for Ti-O parameters, for the triplets Ti-Ti-O I have to use the same cutoff of Ti-Ti-Ti triplets or I have to use a combined cutoff??
-
How can I implement, for Ti-Hf interaction, the following parameters: Addrep, R, D and hfocord. I read the articles about this potential and LAMMPS manual, but I didn’t understand how implement these parameters. Regarding R_omiga_{a,b,c,d}, I would like to be sure that they refers to the first element.
When I succefully implement and test these parameters, I’ll be happy to share them with you.
Thanks in advance
Giulia
M2SCS School of Graduate Studies
University of Modena and Reggio Emilia
Department of Material and Environment Engineering
Strada Vignolese 905,
41125 Modena, Italy
e-mail: giulia.broglia@…3753…
Hi Giulia,
I would like to model the Ti/HfO2 interface, and I have some question about
how to develop the parameters, in particular for Ti-Hf:
- how are the cutoff R and D chosen? Are they pair dependent?For example for
Ti-O parameters, for the triplets Ti-Ti-O I have to use the same cutoff of
Ti-Ti-Ti triplets or I have to use a combined cutoff??
R and D determines the cutoff. R-D is the inner cutoff, R+D is the
outer. Typically inner and outer cutoffs fall between 1st nearest
neighbors (NN) and 2nd NN. Cutoff is only dependent on I-J, so that
cutoffs for Ti-Ti-O should be the same as Ti-Ti-Ti.
- How can I implement, for Ti-Hf interaction, the following parameters:
Addrep, R, D and hfocord. I read the articles about this potential and
LAMMPS manual, but I didn't understand how implement these parameters.
Regarding R_omiga_{a,b,c,d}, I would like to be sure that they refers to the
first element.
Addrep and hfocord: 0.0.
R & D: lorentz berthelot mixing rules, R_inner = sqrt(R_inner_of_I *
R_inner_of_J), same for R_outer. Then convert inner and outer cutoffs
to R & D. If by any chance the cutoffs given by the mixing rule does
not satisfy the 1NN and 2NN condition, you can adjust them to fit
between 1NN and 2NN of Ti-Hf alloy.
R_omega_a,b,c,d are used to scale the A, B, lambda and alpha between
pair interactions. As a first try, you can determine A, B, lambda and
alpha by again the mixing rules and set all R_omega to 1.0. You can
then model Ti-Hf with these parameters, but this set of parameters are
not fitted to any expt or DFT data. It would be best to fit these
parameters to some data.
Best,
Ray