velocity autocorrelation function

dear all users

I am interested in calculating velocity autocorrelation function
of graphene in In contact with water.
I have used lammps-30Jan14 and compute vacf command to obtain VACF of graphene plate in my system. “Fix Ave/time” has been used to output the data. the results from VACF are incorrect, in such a way that the VACF values continuously fluctuate around zero .I do not know what is the problem?
my input script:
fix 1 G spring/self 50.0 xy
fix 4 water shake 0.0001 10 0 m 15.999400 b 3 a 6 #m 15.999400
fix 5 all npt temp 230.0 230.0 100.0 x 1.0 1.0 1000.0 y 1.0 1.0 1000.0

velocity all create 230.0 35281345 rot yes dist gaussian
compute 1 G vacf
fix ct G ave/time 10 1000 10000 c_vac[1] c_vac[2] c_vac[3] c_vac[4] file vacf.txt
variable VACF equal c_vac[4]
fix VAC1 Gl print 100 “${VACF}” file vacprint.txt screen no
thermo 500

the figure attached for VACF data


What does the plot do in the first 10 timesteps?

The VACF does decay to zero, sometimes
rapidly. After 5K steps, I would expect no correlation.