MEAM potential for Al 7075

Dear Lammps Users
I’m trying to to model aluminum 7075 alloy under tensile loading and I’m facing a problem with the pair_coeff command
In the model , I’m considering the following elements (Al, Zn, Mg, Cu, and Si) and I found the following potential files
MgAlZn.library.meam

MgAlZn.parameter.meam

Si_2.meam.spline

library.Cu.meam

Cu.meam

the part of the code regarding the pair_coeff command is as follows

pair_style meam

pair_coeff * * MgAlZn.library.meam Mg Al Zn MgAlZn.parameter.meam Al Zn Mg

pair_style meam/spline

pair_coeff * * Si_2.meam.spline Null Null Null Null Si

​When I run it i get this error “Cannot open MEAM potential file Al”

Can any one please help me to understand why I have this error and plus I’m not sure how to define the potential for Cu

Thanks
Isra Al Muscati

Dear Lammps Users
I’m trying to to model aluminum 7075 alloy under tensile loading and I’m facing a problem with the pair_coeff command

the problem is not with the pair_coeff command, but more conceptually in how you try to set up the model. there are multiple points, some apply to LAMMPS in general, some are specific for MEAM, that you need to take into account

  1. there must be only one pair_style command. if you have a second pair_style command it will wipe out the previous one and all pair_coeff settings associated with that

  2. there is an option, pair_style hybrid, to work around the limitations of 1), but since it is a purely mechanical coupling scheme, it will not work correctly for mixing multiple MEAM or EAM or any other manybody style potentials. it is meant to be used with only pair-wise additive potentials, or when having separate objects, when each individual object may be modeled with a many-body potential, but the inter-object interactions must be modeled with pair-wise additive potentials. using hybrid for multiple MEAM or EAM potentials is particularly bad, since the embedding terms are grossly wrong. all elements would contribute to the embedding term, but with hybrid, each substyle will be treated completely separate, so the contributions to embedding are not included.
    in short, it would be a very bad idea to use pair style hybrid for an alloy.

  3. MEAM has a very particular way of how you specify parameters: a) you have generic per-element parameters that you can extract from a “library” file. These could be merged easily from multiple “library” files. b) from that “library” you extract basic parameters for entries, this is the first set of “elements”. the “library” may have different entries for the same element (just using different names, e.g. Al and AlS). this lookup is independent from the atom types in LAMMPS, c) you can specify a second parameter file, that is specific for the given system, that has parameters (mostly many-body terms) that will override the default many-body terms inferred from the library entries. It is important to note, that this parameter file uses numerical indices (starting from 1) and that those numbers match the index of the elements that are “extracted” from the library. in your example of “pair_coeff * * MgAlZn.library.meam Mg Al Zn MgAlZn.parameter.meam Al Zn Mg” you extract entries “Mg”, “Al”, “Zn” from the library and in MgAlZn.parameter.meam those will be referred to as 1, 2, and 3, respectively. d) the final list maps the extracted elements to the actual atom types used in LAMMPS.

to summarize:

  • you need to create a single library file that has entries for all atom types in your system, but this can be merged from existing MEAM (not meam/spline) library files.
  • you must customize the parameter settings to correctly match the indices due to the order of “extracting” entries from the library file. merging multiple of these parameter files is non-trivial and requires a good understanding of the parameterization methodology and overall structure and nature of the MEAM model.
  • it would be best to find a parameterization that is valid for your specific system. since you will be missing some override settings in the second parameter file with implications to the accuracy of your simulation.
  • there is no automated way where LAMMPS can do the steps of joining the parameter files

… and a final general comment about alloys. most alloys are mixtures where some elements are only present in extremely small percentages. for typical system sizes, leaving those elements out, will usually not cause an error that is significant compared to intrinsic errors from the MEAM model itself and doing atomic scale studies (e.g. finite size effects)

Axel.

Dear Axel
I appreciate your fast response
I have edited the library file by copy pasting the parameters for the needed elements from different library file (as I understood from you)
The problem I have is with parameter file as I’m not sure how to edit it and what should be the format of such a file
I have also tried to run the model by changing the atom types to three and considering only Al, Mg, and Zn( for which I already have their parameter file ) and the run went smooth so the main issue I have now is with the parameter file that takes under consideration all 5 elements of the alloy
can you please explain to me or elaborate on how the format of the parameter file should be
I’m attaching the input file along with the modified library file and the parameter files that I have obtained

Regards
Isra

MgAlZnCuSi.library.meam (1011 Bytes)

MgAlZn.parameter.meam (879 Bytes)

Cu.meam (145 Bytes)

Al_7075_3.in (2.3 KB)

The syntax of the parameter file is documented in the pair style meam/c documentation in the LAMMPS manual and the detailed meaning and derivation of those parameters is given in the cited publications.

As mentioned before, simply merging settings from multiple files will likely lead to an inconsistent or incomplete parameterization, as you are missing “mixed” terms for the specific manybody terms for combinations of atom types that are taken from different parameterizations.

axel.

Dear Axel,
I was able to find the parameters required for the parameter file in the literature and I wrote the parameter file but I am still having an error that says " Did not find all elements in MEAM library file (src/USER-MEAMC/pair_meamc.cpp:478)"

Can you please advise why I am getting this error
Attached is the input file along with the parameter and library files
Thank you
Regards
Isra

Al_7075_4_AlSiMgCu.in (2.3 KB)

AlSiMgCu.parameter.meam (1.28 KB)

AlSiMgCu.library.meam (866 Bytes)