Dear all,
I am going to analyze the diffusion of AlSi12 alloy and Al2O3 (Al2O3/AlSi12 system). I used the following pair_style. I receive no error from Lammps, but I am not sure if my definitions are correct or not.
mass 1 26.981539 # Al (AlSi12)
mass 2 15.999000 # O
mass 3 26.981539 # Al (Al2O3)
mass 4 28.085000 # Si
fix ff1 all property/atom q
pair_style hybrid comb3 polar_off tersoff morse 4 morse 4
pair_coeff * * comb3 ffield.comb3.NiAlO Al O Al NULL
pair_coeff * * tersoff 2007_SiO.tersoff NULL O NULL Si
pair_coeff 1 4 morse 1 0.4824 1.322 2.92
pair_coeff 3 4 morse 2 0.4824 1.322 2.92
Furthermore, I should use charge atom_style for comb3. I don’t know how much charge I should use for the atoms.
In simulations there are no absolutes. Either those are a good model for your system or not so good. Because they are models, there will always be (systematic) errors. The question is, are those errors small enough to not matter for what you want to determine.
That said, if I were a betting person, I would bet a lot of money against this model. The chances are very high that this will produce bogus results. This looks like an input that was put together via cut and paste by somebody that has no understanding of the potential functions that are used and is just mix-and-matching pieces that were used for other systems.
This answer you can answer yourself by studying (understanding) the potential you are using.
In general, the determination of whether some set of potentials are good or bad for a given problem is something that should be at the beginning of every study. In your system with two compounds you would start by studying each compound separately, look up suitable potentials and parameters and reproduce results from the corresponding publications describing the parameters. That will serve as a “ground zero” to start from. For a multi-component system, you need a more complex potential. Ideally, all atom types (not necessarily elements! you may have different atom types for the same element; there is a big difference between metal atoms like aluminium in different oxidation states, for example) are described by the same functional form and the same parameters for the combined system as for the individual systems, but that is not always possible. Then you have to ask yourself whether pair style hybrid is an acceptable choice. From your description of your problem I would guess not. Then you have to keep looking or find somebody that knows how to derive such parameters and is willing to do it for you.
Overall, it very much looks like you are trying to do something for which you do not have the training, i.e. like you want to drive 15 ton truck when you have previously only driven a scooter. Thus my recommendation is: find a sufficiently experienced collaborator!