[lammps-users] Different results : fortran LAMMPS (2001) vs C++ LAMMPS (July 07)

Hi, I have been using the old version of LAMMPS for the past couple of years and decided recently to change to the current version. I did tests on LJ polymers with degree of polymerization, N =200, bonded using FENE bonds. And I used a cosine angle bending potential to make the chain semiflexible. I ran the two systems in a 200 x 200 x 200 sigma simulation box using a langevin thermostat fix. ( I attached both old and new lammps input files). What puzzles me is that i believe that i translated each command from 2001 lammps to c++ lammps correctly however the <R^2(n)> vs n plot is different for each system.( <R^2(n)> is the averaged square distance between n number of beads). Please can you take a look at both input files and tell me where I went wrong in the translation to current version. I also attached the data file for reference

in.chain-oldlammps (1.04 KB)

data1chain-oldlammps.dat (14.1 KB)

data1chain-newlammps.dat (14.1 KB)

in.chain-newlammps (1.08 KB)

oldvsnewlammps.jpg