ERROR: Incorrect args for pair coefficients

Dear Sirs

Thanks for your good advice. I solved the previous problem. I found
one atom was overlapping. Now it is ok. But another problem is......
I am facing ERROR: Incorrect args for pair coefficients
(../pair_eam.cpp:389) I have 3 types atoms. 1,2 and 3. Type 1 is
graphite, 2 & 3 are Cu.
Type 3 is fixed bottom part of substrate. Would you please tell me
exactly where I am doing wrong?and for interaction between Cu and
graphite I am using
lj potential. Still now I don't have any referance for that potential.
I badly need is my input file

# LAMMPS Input File

# ---------- Initialize Simulation ---------------------
units metal
dimension 3
boundary p p f
atom_style atomic

# ---------- Create Atomistic Structure ---------------------
lattice fcc 1.0
read_data Cu_graph.dat

group graphite type 1
group substrate type 2
group bottom type 3

fix b2 bottom setforce 0.0 0.0 0.0

# ---------- Define Interatomic Potential ---------------------
pair_style hybrid eam lj/cut 4.0 airebo 3.0 0 0
pair_coeff * * airebo CH.airebo NULL C NULL
pair_coeff 2 2 eam Cu_u3.eam
pair_coeff 3 3 eam Cu_u3.eam
pair_coeff 1 2 lj/cut 0.0042038 3.37
pair_coeff 1 3 lj/cut 0.0042038 3.37
pair_coeff 2 3 eam Cu_u3.eam

neighbor 2.0 bin
neigh_modify delay 10 check yes

# ---------- Run Minimization ---------------------
reset_timestep 0
timestep 0.001
thermo 10
thermo_style custom step temp press pe ke etotal enthalpy
minimize 1e-10 1e-10 1000 1000

dump 1 all xyz 500

velocity graphite set 0.0 0.0 -3.2

fix 10 all nve
run 20000


You need to use “pair_coeff * * …” for eam. Please have a look at eam’s doc page for more info.


You need to use "pair_coeff * * ..." for eam. Please have a look at eam's
doc page for more info.

sorry ray, but that is not the correct answer. unlike the other eam
variants, plain eam *does* support numerical values, but it follows
the same restrictions from other manybody potentials, i.e. it cannot
be used as a pairwise additive potential, but has to include atom
types wholly.

running the input with the -echo screen flag quickly reveals that the
following line is the offender

pair_coeff 2 3 eam Cu_u3.eam

which indeed violates the requirements for manybody potentials. but
since this line is part of a whole set of three pair_coeff statements,
it can be solved in a rather simple fashion: all three lines can be
replaced by one single pair_coeff statement that does not violate the

pair_coeff 2*3 2*3 eam Cu_u3.eam

this will cover 2 2, 3 3 *and* 2 3.


Thanks, Axel! Learnt one more thing.