Hi Steve,
I uploaded to the end my input file. I am monitoring temperature from output file so I guess it is entire system temperature. When I monitor via graphical interface, again I see atoms are not moving in parallel to output freezing point. I can try dumping velocity before that you can take a look at my input.
units metal
atom_style charge
boundary p p p
read_data ets-10-large-box2.lj
velocity all create 300 99999999 dist gaussian
pair_style hybrid comb coul/cut 10.0
pair_coeff * * comb ffield.comb NULL Si O NULL
pair_coeff * * comb ffield.comb Ti NULL O NULL
pair_coeff 1 1 coul/cut 20.0
pair_coeff 2 2 coul/cut 20.0
pair_coeff 3 3 coul/cut 20.0
pair_coeff 4 4 coul/cut 20.0
pair_coeff 1 2 coul/cut 20.0
pair_coeff 1 4 coul/cut 20.0
pair_coeff 2 4 coul/cut 20.0
pair_coeff 3 4 coul/cut 20.0
neighbor 0.5 bin
neigh_modify every 1 delay 0 check no page 100000 one 10000
fix 1 all nvt temp 300 310 0.0001
dump 1 all xyz 200 combTiSi300to310K04092.xyz
dump_modify 1 append yes
thermo 50
timestep 0.0001
run 4000
Mehmet KOC